increase testosterone is to reduce estrogen. The most effective way to do this is maintain a low level of bodyfat.

Adopt a testosterone-enhancing diet

cruciferous vegetables such as broccoli, cauliflower, kale, swiss chard, brussels sprouts and cabbage. These vegetables contain phytochemicals, which are essential for healthy hormone production and estrogen metabolism.Celery 

Limit excessive carbohydrate intake

testosterone, and all other male hormones, are actually made from cholesterol.

Eat many smaller meals: eliminate unnecessary stress hormone release. Don’t wait until you are hungry to eat.
Vitamin A: 25,000 IU 
Vitamin E: 400 IU daily 
To reduce free radical stress on the pituitary gland take an extra 400IU prior to a work-ou
Vitamin C
Vitamin B Complex 
Zinc
Boron
Selenium
1.Glycine: Glycine has been found to bind with certain toxins so they can be safely excreted. Luckily, glycine is cheap, and a gram a day would be enough for most people.


2.Vitamin C: Helps build glutathione and protects against phthalates. Take 3 grams a day. However, the best way to determine vitamin C dose is bowel tolerance.


3.Glutamine: Helps build the gut. The healthier the gut, the more it can protect you against incoming toxins. If you have leaky gut syndrome, you could take 5 to 10 grams with each meal or 60 to 80 grams a day. As for a maintenance dose, 20 to 30 grams should do the job.


4.Taurine: This is a calming amino acid but also has the ability to activate detoxification pathways. Take this either after a workout or before bed. One to three grams is best.


5.Zinc: A critical mineral for testosterone. In fact, 98% of zinc in males is stored in the prostate. Low zinc status can and will affect testosterone production. As for dosing, try 300 mgs per day. Some may benefit from more, especially if they have high copper.


6.Antioxidant supplements high in A, E, and D: My favourite way to hit this is to simply use organic butter.


7.Resveratrol: This simple polyphenol is anti-estrogenic.

Daily physical activity: Testosterone levels increase most with short intense bursts of activity. They decrease with prolonged endurance activities 
forty-five minutes per session 

sleep deprived, your testosterone will suffer. 

sunlight: one hour of exposure 

Increase your stress resistance: There is a strong relationship between mental outlook and physical well-being and it is largely controlled by hormones. While muscular stress is very beneficial to testosterone, the wrong kind can be devastating. Emotional stress is a frequent cause of decreased testosterone levels. Chronic stress not only interferes with testosterone function, muscle building and strength, it causes premature aging and contributes to the onset of cardiovascular disease.

ENVIRONMENTAL TOXINS that decrease test

 Bisphenol A (BPA): the skin does a great job of absorbing it. The hormone BPA best represents is a nasty form of estrogen. Dosing with smaller amounts just doesn't work for BPA.
BPA acts as an estrogen, but with two differences:

1.It's foreign to the body.
2.It's more harmful than natural estrogens.
Xeno-estrogens do all the things that we don't want. Think of it as the complete opposite of injecting testosterone. Xeno-estrogens decrease testosterone and increase estrogens
To know if BPA is in your bottle, look for a triangle with a 3 or a 7

Phthalates
Phthalates can be used in virtually anything from your girlfriend's sex toys to your sex toys. 

Seriously, phthalates are used in everything: air-fresheners, cosmetics, shampoos, children's toys, and paints. Why the heck is a plasticiser used in air-fresheners and things that smell nice? 

 Phthalates hold aromas. So that car-freshener you use to hide the smell of your farts is also lowering your testosterone. (Maybe you should just lower the window from now on.)

Phthalates, like BPA, suppress testosterone, increase insulin resistance, and chelate magnesium and zinc. (10, 11) Their impact on zinc and magnesium can have a very negative compounding effect.

An interesting correlation I learned from Mark Schauss, author of Achieving Victory Over a Toxic World, is that the explosion of autism occurred simultaneously with the introduction of phthalates in 1970. Dr. Schauss would be the first person to point out that it's not only the phthalates, but also the negative synergetic effect from the mass amount of environmental toxins. 

HomeProtein & Amino AcidsDietary Fats & CarbohydratesVitamins & MineralsAnti-Oxidants & Anti-NutrientsStrength & ConditioningPhysiology & MetabolismHormonesHumourLinksContactVulgar Adverts

Raise Testosterone Naturally!
September 2010
Revised: December 2010 and January 2011


1.Introduction
2.High- and Low-Fat, High- and Low-Carbohydrate and High- and Low-Fibre Diets
3."In Vitro" Experiments on Fatty Acids, Sex Hormones and Sperm Cells
4.Negative Energy Balance, Fasting and Starvation
5.Dietary Protein and Supplemental Amino Acids, BCAAs, Creatine etc.
6.Alcohol (Ethanol), Testosterone Levels and Testicular Functions
7.Vitamin A (Retinol/Retinal, Retinoic Acid and Carotenoids)
8.Vitamin D
9.Vitamin K
10.Vitamin E (Tocopherols and Tocotrienols)
11.Vitamin C (Ascorbic Acid), Glutathione and Uric Acid
12.B Vitamins
13.Macro-Minerals (Sodium, Potassium, Phosphorus, Calcium, Magnesium) and Boron
14.Trace Minerals (e.g. Iodine), Goitrogens and Toxic Metals and Ions (e.g. Fluoride)
15.Zinc
16.Garlic, Onion and Allium Plants Containing Diallyl Sulfide
17.Fenugreek
18.Mucuna pruriens (Velvet Bean)
19.Serenoa repens (Saw Palmetto or Permixon)
20.Tribulus Terrestris
21.Stinging Nettle
22.Mint (Mentha) and Other Lamiaceae Plants (e.g. Basil, Thyme, Sage, Marjoram)
23.Soy Products and Isoflavones
24.Liquorice, Liquorice Root and Glycyrrhizin
25.Methylxanthine Alkaloids, Theobromine, Theophylline, Caffeine, Coffee, Tea, Cocoa and Chocolate etc.
26.Other Alkaloids, Nicotine and Tobacco Smoke, and Recreational Drugs
27.Strength, Power and "Anaerobic" Training
28."Aerobic" Endurance Training
29.Stress Steroids, CNS Aromatase and Serotonin, Impulsive Aggression and Adaptive Insulin Resistance
30.Music, Films, TV, Games, Sports and Miscellaneous Entertainment
31.Tip-Top Tips to Top-Up Your Testosterone Tank

Introduction
To judge from its title, this article is about a single hormone and how to go about getting more of it. Actually, it's far more complicated than that, but a title representing its full scope would attract minimal attention. Before I set forth, here is some background information through which the contents of this article need to be considered.

Testosterone, produced via the sexual organs or the adrenal glands, is popularly regarded as being the ultimate promoter of "maleness" (whether that is depicted as something strong, muscular and courageous or as something selfish, violent and anti-social) and as the ultimate "anabolic", growth-promoting entity (whether the growth medium is skeletal muscle, internal organ, bone tissue or prostate cancer cell). It is also regarded as the ultimate marker of male sexual potency.

It can be metabolized (via the enzyme 5-alpha reductase) into dihydrotestosterone (DHT). DHT is present in the human body in much smaller quantities than testosterone, but it is much more potent than testosterone at doing some of what testosterone does, as well as being widely regarded (in some quarters) as being the most potent promotor of prostate cancer progression. Fearing the amount of extra research to be done, I initially decided to avoid any talk of the relationship between steroid hormones and prostate cancer, but I changed my mind after learning some fascinating things involving vitamin D3 (see the section dealing with vitamin D for further details).

Most research concerning "male" steroid hormones focuses solely on testosterone levels. All else being equal, a person producing more testosterone will also be producing more DHT (and other androgens, sex hormones, steroid hormones and related entities), but all else rarely is equal. Other important things include bioavailable testosterone and DHT, "female" hormones (present in similar quantities in men), stress hormones, hormone receptors, "gonadotropins" and the amount of testosterone and DHT actually present in the semen rather than just the serum.

Most of the testosterone (and DHT and estradiol) in the human body is bound by the aptly named "sex hormone binding globulin" (SHBG), a globulin which binds sex hormones. Sex hormones bound by SHBG are regarded as being "inactive", but being bound by it does not necessarily mean that they cannot be transported by it to a cell (or the CNS) and there utilized or in some way responded to by other chemical entities (i.e. made active to some extent). A high level of total testosterone ("free" or bound) is not to be scorned, but SHBG appears to have a diminished affinity for its cellular receptors when it is already "liganded" by binding a steroid hormone, especially when DHT is the ligand. If it has already bound to a receptor in a non-liganded state, SHBG can then efficiently bind sex hormones which potentially can exert some effect in that locale. Steroid hormones are able to interact with cells via SHBG receptors in addition to their own, in a process that generates cyclic adenosine monophosphate (cAMP) as a secondary messenger. Thus, non-SHBG-bound testosterone does appear to be rightfully regarded as an important factor, but non-steroid-bound SHBG appears to have uses of its own. Instead of being bound to SHBG, some testosterone can be bound by an albumin-based "androgen-binding protein" (ABP), an androgen binder which is considered to be at least partly "bioavailable". In an experiment on adult male Sprague-Dawley rats, SHBG (not normally present in rats) delayed the transport of testosterone from the serum into the CNS cerebrospinal fluid, whereas ABP (normally present to some degree) had no such effect. Some research regarding "male" hormones combines "free" testosterone and albumin-bound testosterone in a measurement known as the "free androgen index" (FAI).

Although testosterone is regarded as a "male" hormone (an androgen) and estradiol (the main estrogen) as a "female" hormone, it's far more complex than that. Testosterone is the metabolic precursor to the estrogens via the aromatase enzyme (one of its own precursors, androstenedione, can also be metabolized directly into estrone and then estradiol via the same enzyme), and aromatized derivatives of androstenedione/testosterone make an important contribution to bone growth and preservation in men, as does DHT. Without being aromatized into estrogens or reduced into DHT, testosterone itself seems to have little effect on bone. Having robust hip bones helps women to bear children, but it can also help men (and women) to be better at squats and deadlifts (which is obviously a far more important consideration from an evolutionary point of view, since, as Jon-Pall Sigmarsson once said, there is no point in being alive if you can't do the deadlift). Estrogens may encourage aggressive/protective behaviour, bone growth or subcutaneous fat deposition depending on their location, the density of their receptors and the receptors' affinity for them.

"Free" testosterone is thought to promote "male" behaviour, and so it does, but this is achieved not only by activating central nervous system (CNS) androgen receptors (and estrogen receptors) but also by being converted into estradiol before or after crossing the blood-brain barrier and entering the CNS. Either hormone can cross the blood-brain barrier and promote certain varieties of "male" behaviour in various regions, but the particular regions affected and the way in which they are affected will depend on how the CNS has been structured in response to a surge of testosterone-derived estradiol during a phase of fetal development, and on the subsequent activation of androgen receptors throughout post-fetal life by testosterone and DHT. Minus fetal estradiol exposure and/or later androgen receptor activation, the CNS is considered to be inherently "female". DHT is not aromatized into estradiol and encourages certain very "male" physical and mental attributes, but estradiol also encourages some "male" behaviours (including some separate ones) when its impact is sufficiently moderated by the activation of androgen receptors (by testosterone or DHT) and by testosterone seemingly "co-localizing" androgen concentrations with the activation of estrogen receptors and "co-localizing" aromatase activity (and thus estradiol concentrations) with the activation of androgen receptors. Whether an influx of estradiol promotes "male" or "female" behaviour in a certain brain region may depend on the extent of androgen receptor activation therein.

Estradiol is characterized as "de-feminizing" the CNS during fetal development; androgens (especially DHT) are characterized as "masculinizing" the CNS (a distinct concept). In rodents and other species with short gestation periods, on whom much of the research concerning the effect of sex hormones on the CNS has been conducted, estradiol has a more extensive role in promoting "masculinized" behaviour, although it certainly also promotes "feminized" behaviour in female rodents (and in male ones with non-functional androgen receptors). In humans and many other primates, however, the activation of androgen receptors is of equal or probably greater importance. DHT activates androgen receptors more potently (unit for unit) than testosterone, but the overall impact of testosterone (even unit for unit) may be of equal or greater importance, since it appears to increase the expression of androgen receptors as a result of activating estrogen receptors (something that DHT cannot do), and also to increase aromatase activity (the estradiol from which is likely to promote male-typical behaviour under the circumstances) as a result of activating androgen receptors.

Males lacking 5-alpha reductase activity due to an absence of the type 2 isoenzyme typically sport ambiguous external genitalia (pseudo-hermaphroditism) in early life, but a surge of gonadotropins and testosterone starting in puberty produces a phenotype characterized by less (if any) body hair, deep voices, muscularity and male-typical identities and sexual behaviours. Males lacking the aromatase enzyme typically (given fetal exposure to estradiol) develop male-typical physical features, identities and sexual behaviours from the off, but case studies suggest that exogenous estradiol matures bones (epiphyseal closure) and increases lumbar spine bone mineral density, and diminishes depression and anxiety while increasing aspects of sexual motivation including libido, erotic fantasies and orgasms (findings that are reinforced by experiments on aromatase knockout male mice). Both androgens and estradiol seem to be important for maximizing sexual desire in both sexes, since pre-menopausal and post-menopausal females reporting diminished libidos often prove to have very low sex-typical levels of free testosterone and/or adrenal-produced DHEA-sulfate (DHEA-S, immediate precursor to the androgens).

All sex hormones are "steroids" but not all "steroid" hormones are sex hormones. While testosterone is an "anabolic" (building up) steroid hormone, cortisol (a "corticosteroid") is regarded as being a "catabolic" (breaking down) steroid hormone. Both hormones are produced via the adrenal gland as part of the so-called "fight or flight" response (the non-steroid hormones noradrenaline, adrenaline and adrenocorticotropin/ACTH are also heavily involved), with cortisol encouraging more "flight" and testosterone more "fight". Like testosterone, cortisol can be "free" or bound, and the "free" versions of each are often estimated (seemingly with great accuracy) by "salivary" measurements.

Cortisol is often contrasted as "bad" compared to testosterone's "good", but cortisol is present in far higher quantities than testosterone and must serve some vital functions. Left to its own devices, it does inhibit (via peripheral insulin resistance) the storage of glycogen in skeletal muscles, and it inhibits bone and muscle growth, but these things depend on the ratio of cortisol to testosterone and are usually just a transient response to a physical challenge. This response prevents runaway hypoglycemia and makes more glucose and fatty acids available for immediate use. Existing glycogen stores can fuel the muscles during physical challenges; dwindling stores can be replenished (especially during and after physical exertion) by mechanisms independent of insulin, and testosterone (the other factor) promotes peripheral insulin sensitivity and thus provides balance.

Interestingly, both cortisol and testosterone are considered to be "immunosuppressive", with high testosterone levels supposedly being a way for males of a great variety of species to show off how superb their immune systems must be for them to tolerate such high concentrations of an "immunosuppressive" substance, but it could be that both hormones actually "redistribute" the immune system to locations on the body that are more relevant to the challenges likely to be faced by those producing them, but less likely to be sampled by researchers.

Testosterone and particularly its "free" form is popularly blamed for various aggressive behavioural displays that are considered to be anti-social. It seems that it is actually CNS aromatase activity producing estradiol that promotes aggression, as well as paternal protection. Free testosterone levels have been associated with "anti-social" behaviour in some cases, but in other cases "anti-social" males have been characterized by high levels of SHBG. High levels of cortisol (the other "fight or flight" steroid hormone) can increase the binding of testosterone by SHBG (and possibly by corticosteroid-binding globulin, CBG, as well), and cortisol also has been linked (when co-elevated with testosterone as part of an adrenal stress response) with the incidence of aggressive behaviour by males (despite also being linked with "submissive" behaviour when testosterone is depleted). Consequently, almost any steroid hormone profile could be linked with "anti-social" behaviour, and the dominant impression emerging from this research is that men can sometimes be aggressive. Who would have guessed? (I have included a less flippant section on this topic near the end.)

Testosterone, aromatase, DHT, "free" or "active" versus bound androgens, liganded versus non-liganded SHBG and the ratio of androgens to cortisol are merely part of the story. Other things that need to be considered are the number of androgen receptors, their location and their affinity for various steroid hormones. Theoretically, a person could have very high levels of androgens but a lack of androgen receptors, or lots of receptors with minimal affinity.

Another very important thing to consider is sex drive and sperm function. High testosterone levels usually correlate with a high sex drive and efficient sperm function, but there are examples of substances and factors that produce higher testosterone levels (in the serum but not necessarily in vital cells) while also messing up testicular function (as mentioned in various sections). Beware!

Besides testosterone, there are some important non-steroidal "gonadotropins", particularly follicle-stimulating hormone (FSH, which stimulates the testicular Sertoli cells) and luteinizing hormone (LH, which stimulates the testicular Leydig cells). Some things that raise testosterone levels do so by increasing levels of these "helpers", but excessive secretion of them can lower testosterone via "negative feedback". High levels of FSH and/or LH are sometimes an indication that the body is trying to respond to diminished testosterone levels. There is another gonadotropin, chorionic gonadotropin, but in humans this occurs in significant amounts only in pregnant women, Brian Cushing and people with certain tumors - including pancreatic islet cell and adrenal carcinomas, which, as irony would have it, are sometimes a feature of the condition known as Cushing's Syndrome!

Here is a list of abbreviations for various steroid hormones, their agonists and transporters:

DHEA = dehydroepiandrosterone
DHEA-S = dehydroepiandrosterone sulfate
DHT = dihydrotestosterone
SHBG = sex hormone binding globulin
CBG = corticosteroid binding globulin
LH = luteinizing hormone
FSH = follicle-stimulating hormone
hCG = human chorionic gonadotropin

Steroid hormone receptors belong to a family of nuclear receptors that also includes the vitamin D receptor, the retinoid X receptor and retinoic acid receptor (concerned with vitamin A and its by-products), and the thyroid hormone receptors.

Like vitamin D3 (cholecalciferol), but in contrast to vitamin A and the thyroid hormones (T4 and T3, derived from the mineral iodine and the amino acid tyrosine), the steroid hormones are derived metabolically from cholesterol, which is itself derived from a substance called squalene that serves as the precursor to other, similar sterols and/or hormones produced by plants and microbes. The cholesterol-to-hormones steroidogenesis pathway eventually reaches pregnenolone, which can lead to progesterone and then the stress or sex steroids, or to dehydroepiandrosterone (DHEA) and then the androgens and estrogens.

In the absence of dietary cholesterol, human beings (except for those suffering from some serious health conditions) are perfectly capable of manufacturing their own cholesterol via other substances in the diet, a process that is especially vigorous in response to palmitic acid (and many other saturated fatty acids) and fructose (and other sugars). Cholesterol is transported through the bloodstream in lipoproteins, including the famed and misunderstood low-density lipoproteins (LDLs) and high-density lipoproteins (HDLs). Both LDLs and HDLs can be a site for steroid hormone production (e.g. in the Leydig cells of live rats), but HDLs and especially HDL3s seem to be a particularly popular resort. HDL-cholesterol levels have been correlated with total and salivary testosterone (but not SHBG) levels among middle-aged men, but were negatively correlated with testosterone levels and positively correlated with SHBG levels among Mormons and vegetarian Seventh-Day Adventists.

High- and Low-Fat, High- and Low-Carbohydrate and High- and Low-Fibre Diets
Several feeding experiments (typically "semi-controlled", in which all food has been provided to subjects who then go off to do their own thing without being monitored to ensure full compliance) have recorded the effect of diets varying in total fat content, the ratio of saturated to polyunsaturated fatty acids (PUFAs) and the amount of fibre. From these it is fairly clear that very-low-fat diets typically lower various measures of androgen levels in men, but the independent effects of the three factors mentioned above have not been thoroughgoingly gauged in live human males.

In one example, males aged 50-60 (an age by which androgen levels are typically lowering) were fed a diet containing 65-70% carbohydrate (types and foods not mentioned), under 15% fat (with unchanged ratios of fatty acid classes) and 25-30g of fiber per day (judged to be "high", although it is about the same as the "low" amount fed in another experiment cited below) for eight weeks in comparison to baseline and/or crossover diets that were deemed to be "high-fat" (more than 30% of calories ... gasp) and low-fibre (under 20g per day). As a result of the low-fat diet, their DHT, total and free testosterone, dehydroepiandrosterone sulfate (DHEA-S) and androstenedione levels all dropped very markedly, and SHBG, estradiol, FSH and luteinizing hormone all dropped to lesser degrees. It is important to distinguish between measures of SHBG itself and measures of the amount of sex hormones bound by it. Whereas hormone-binding SHBG has diminished affinity for cellular receptors, non-"liganded" SHBG has greater affinity for cellular receptors and is able to activate them (and subsequenty bind hormones) once received. In this study, the low-fat diet lowered SHBG itself. The intervention was lower in fat and higher in fibre, but it seems likely that reducing fat from over 100g to 40g had a much greater impact than increasing fibre by a measly 10g or so.

In another example, with middle-aged men whose exact ages are not given in the abstract, a lower-fat and higher-fibre diet along with an increased ratio of polyunsaturated to saturated fatty acids resulted in reduced total and free testosterone levels that returned to normal after resumption of the baseline diet.

Another experiment (with men whose ages are not specified in the abstract) tested the combined effect of reducing total fat (from 40% to 25%) and increasing the ratio of PUFAs to saturated fatty acids (from 0.15 to 1.22) while (so it seems) keeping every other aspect of the diet (including fibre) fixed. The result was that total and free testosterone dropped significantly and estradiol dropped non-significantly.

Not surprisingly, a group of men whose fat intake was reduced from over 100g per day to under 20g per day exhibited lower free testosterone levels and (in conjunction with that) higher SHBG levels.

Among men aged 19-56, 10-week "high-fat", "low-fibre" (41% fat versus 45% carbohydrate and c. 26g fibre per day) and low-fat, high-fibre (19% fat versus 67% carbohydrate and c. 61g fibre per day) diets (separated by two-week washout phases) had a minor effect on hormone levels (total testosterone was not-quite-significantly higher on the higher-fat diet, a difference explained by proportionally higher levels of the SHBG-bound fraction). Urinary excretion of testosterone was higher on the high-fat, low-fibre diet. Perhaps an important difference between this study and the first one cited above is that it also contained younger subjects who were probably better able to metabolize a high carbohydrate intake.

Athletes are also better able to metabolize a high carbohydrate intake. Among elite male ice hockey players, a 10% increase of carbohydrate intake (to 55%) at the expense of fat resulted in increases in both non-SHBG testosterone and cortisol (a measure of anabolism versus catabolism), the ratio of which remained the same.

Among pre-menopausal women, two months of a low-fat diet (20% of calories, types unspecified in the abstract) reduced non-SHBG estradiol and testosterone compared to a crossover higher-fat diet (40%). Hormones were altered little in post-menopausal women.

A range of experiments (8-10 weeks over two menstrual cycles) in other pre-menopausal women found that, compared to "high-fat" (i.e. moderate-fat, 40%) and low-fibre (12g per day) diets, diets low in fat (20-25%) and higher in fibre (40g per day) lowered levels of estrone, estrone sulfate, testosterone and androstenedione, and not-quite-significantly lowered free estradiol. The increased fibre intake was judged to have independently lowered both total (but not free) estradiol and SHBG (which probably represents a mild diminishment of overall estrogenicity).

Pre-pubescent girls aged 8-10 were pressured via "behaviourists" and American Heart Association "educational" materials to eat loads of fibre, to restrict intake of cholesterol (precursor to the sex hormones) to 150mg per day at the absolute most, and to limit fat to 28% of calories with PUFAs providing slightly more of the fat than saturates. The end result of this wonderful intervention/abuse of 7-9 years was that total and free estradiol (and follicular phase estrone sulfate) were lower by 30% compared to the control group, while luteal phase progesterone was more than 50% lower (possibly indicative of anovulatory cycles) and luteal phase testosterone nearly 30% higher (due to presumed diminished aromatase activity).

Unusually high testosterone levels and anovulatory cycles are often a feature of polycystic ovary syndrome (PCOS) in women. When women with PCOS were fed single meals that were low in fat and high in carbohydrates and fibre (6% fat, 81% carbs and 27g of fibre) or a "high-fat western meal" (actually bearing no resemblance to typical western meals, with 62% fat, 24% carbs and 1g of fibre), testosterone levels initially dropped to the same degree after both meals but remained lower for longer after the high-fat meal. Estradiol levels were not measured. Glucose and insulin levels (another feature of PCOS) were higher after the low-fat diet. DHEA-S (initially before rising) and cortisol (throughout) were lowered after both meals.

Taking all of the above into consideration, it's clear enough that (at least in typical research subjects) low- to very-low-fat diets reduce testosterone levels whereas higher-fat diets appear to increase levels of sex hormones and SHBG across the board. Higher fibre intakes have always accompanied low or very low fat intakes and have been found guilty by association, but it is possible that they have little or no effect on the "free" concentrations of the sex hormones and that they mainly lower levels of total testosterone (or estradiol in women) and SHBG (which probably represents a mild anti-androgenic/estrogenic impact). A high ratio of PUFAs to saturated fatty acids has always accompanied reduced total fat intakes, and only a few experiments have included this factor, but it has always been associated with reduced free concentrations of testosterone or estradiol (depending on the sex of the subjects).

Since low-fat diets typically lower testosterone levels, does that mean that the best way to raise testosterone is by eating higher and higher proportions of fat? Not likely. It is important to distinguish between the secretly-quite-high-carb "high-fat western" diet used by most researchers and diets in which the calories actually are dominated by fat. Unfortunately, research regarding truly high-fat diets and sex hormones is thin on the ground and based on very short term observations. In one example, a single high-fat meal (% and type not specified in the abstract) lowered both total and free testosterone in "normal" men, unlike a non-nutritive meal and a meal with minimal fat and mixed protein and carbohydrates. In another example, men who had followed a very-high-fat, very-low-carb diet (66% versus 4%) for four weeks showed (by comparison with when eating equal portions of fat and carbohydrate) higher cortisol concentrations and a faster appearance of cortisol and a delayed clearance of cortisol metabolites in response to a cortisol infusion. A three-day very-low-carb diet (50% fat, 45% protein, 5% carbs) did not affect testosterone responses to incremental exercise in a group of men, but it did reduce basal testosterone production. Lower levels of free testosterone were seen for up to eight hours after 86%-fat meals in males who had followed a 64%-fat diet for eight weeks, although the eight-week diet did not itself affect any hormone measure and it was speculated (among other possibilities including PUFAs diminishing the binding capacity of the luteinizing hormone receptor in the absence of high cholesterol intakes) that higher cellular uptake of testosterone was responsible for the post-prandial reductions. The 86%-fat meal contained 52g of saturated fat, 59g of MUFAs mostly from olive oil and 12g of PUFAs mostly from fish oil. I suspect, however, that a very pronounced shortage of carbohydrate (such as is inevitable on an 86%-fat diet but not for many people on a 64%-fat diet) will inevitably produce a bit of metabolic stress. It appears that both an over-abundance and an under-abundance of carbohydrate (relative to the individual's carbohydrate turnover) has the potential to adversely affect the hormonal environment.

As for specific foods, specifically meat, one observational study observed higher total but similar free testosterone levels in meat-eating males compared to lacto-ovo vegetarian ones (who presumably were not adhering to the "hormone precursor diet"), and another found higher total but similar free testosterone levels (along with lower insulin-like growth factor I levels) in vegan males (compared to meat-eaters and lacto-ovo vegetarians) whose diets tended to be low in fat with a high ratio of PUFAs to saturated fatty acids (the vegans were about 10 years younger and 10kg lighter on average). A single-meal study found lower post-prandial testosterone levels for up to six hours after a meal of lean meat (see also the protein section), and testosterone was kept at basal levels by adding animal fat to the meals (but not by adding safflower oil, rich in polyunsaturated linoleic acid).

The totality of the above research doesn't leave things crystal clear. The rest of this section is my interpretation of it in the light of my working knowledge and theories.

Since cholesterol is the precursor to the steroid and sex hormones, higher cholesterol intakes will, all else being equal, lead to greater production of these hormones than lower cholesterol intakes. However, there is very likely to be a point at which higher hormone production from higher cholesterol intakes reaches a ceiling (the thyroid gland plays a very important role in metabolizing cholesterol, and the most active thyroid hormone is able to lower testosterone production via "negative feedback"). Since dietary sources of cholesterol also contain animal fats rich in palmitic and stearic (saturated) and oleic (monounsaturated) acids, lowering intake of animal fat is likely to lower intake of cholesterol as well, meaning that the link between low-fat diets and lower testosterone levels could be partly confounded by lower cholesterol intakes.

When there is less cholesterol available, the body is quite capable of synthesizing its own, notably when there is a generous supply of saturated fatty acids with 12-16 carbon chains (lauric, myristic and palmitic acid). These saturated fatty acids tend to increase levels of both low- and high-density lipoproteins (LDLs and HDLs) compared to most other dietary components, and cholesterol from which hormones can be synthesized is transported around the body in these. HDLs seem to be particularly associated with androgens, and monounsaturated oleic acid (but not monounsaturated palmitoleic acid) also tends to increase levels of these.

When fat as well as cholesterol is in short supply, the body is quite capable of synthesizing both via carbohydrate intake, and a little known fact is that (of all dietary components) sugars such as fructose and sucrose (a mix of fructose and glucose atoms) tend to raise LDL levels the most. If the skeletal muscles are not sensitive to insulin, glucose and possibly even starch will do this too. Despite the theoretical ability of high carbohydrate intakes to increase production of cholesterol-derived hormones, sedentary populations tend to overeat and become hyperglycemic, glucose intolerant, hyperinsulinemic, insulin resistant, viscerally obese and potentially diabetic on such diets. This typically results (in type-2 diabetic women) in higher testosterone but lower estradiol levels and (in type-2 diabetic men) lower total but comparable free testosterone levels - the kind of profile that seems to be encouraged by higher fibre intakes. Males with type-1 diabetes sport higher SHBG levels in conjuction with their lack of endogenous insulin, whereas a non-significant correlation between insulin resistance and higher free testosterone has been detected among their non-diabetic siblings.

Interestingly, the process that can lead to the unbalanced accumulation of fat around the internal organs behind the abdominal wall, "de novo lipogenesis" (which can be enhanced by ethanol, soluble fibre and medium-chain fatty acids as well as by high carbohydrate intakes), may also serve to increase levels of free testosterone (if mild and acute rather than severe and chronic). This could well explain why some high-level athletes (whose carbohydrate turnover is much greater) excel at testosterone-associated endeavours on very-high-carbohydrate diets, and why the elite ice hockey players from one of the experiments cited above (in contrast to the members of the general public from other experiments) exhibited higher free testosterone concentrations when they increased carbohydrate intake by 10% at the expense of fat. For the vast majority of people in the vast majority of situations, fairly even intakes of fat and carbohydrate are very unlikely to lower sex hormone production.

"In vitro" experiments using cultured mouse livers support the notion that sugars, especially fructose, are able to diminish hepatic production of SHBG (not desirable in and of itself) through the mechanism of "de novo lipogenesis", which involves the production of cholesterol (precursor to the sex hormones) and palmitic acid via acetyl-CoA. This mechanism is also linked with satiety from feeding, and with transient, mild insulin resistance and increased insulin output, but chronic insulin resistance and hyperinsulinemia have an adverse effect on sex hormone production (hence the low SHBG and total testosterone but normal free testosterone levels reported in diabetic males).

It is not really accurate to say that sex hormone production is "increased" by a diet that is not low in animal fats nor cholesterol and neither too high nor too low in carbohydrate. Rather, such a diet enables people to produce the amount of sex hormones that they'd be producing if they weren't being dickheads and taking advice from the American Heart Association or Robert Atkins. 

"In Vitro" Experiments on Fatty Acids, Sex Hormones and Sperm Cells
Some over-hyped herbal "testosterone boosters" are also marketed as "prostate-friendly" supplements because their unusual free fatty acid contents (fatty acids in foods are usually "esterified" in triglycerides and thus not "free") will supposedly inhibit 5-alpha reductase activity and thus raise testosterone but lower DHT. See the sections on Serenoa repens and Tribulus terrestris for a summary of their lousy track record as supplements (including case studies involving pancreatitis and hepatitis). In fact, "in vitro" experiments do indicate that most free fatty acids inhibit 5-alpha reductase. The catches are that those fatty acids can be rapidly "esterified" in an "in vivo" environment (that of a living organism, e.g. a human being) and that not all free fatty acids inhibit 5-alpha reductase and DHT production even in an "in vitro" environment.

In an "in vitro" experiment using fatty acids from permixon (Serenoa repens), about 80% of which are free fatty acids, no esterified fatty acids were seen to inhibit 5-alpha reductase activity. As a free fatty acid, lauric acid (a saturated fatty acid with 12 carbon chains, found in large quantities in coconut oil) inhibited both the type 1 and the type 2 isoenzyme of 5-alpha reductase. Free myristic acid (a saturated fatty acid with 14 carbon chains, found in significant quantities in dairy fat) inhibited the type 2 isoenzyme but (for some reason) was not tested on the type 1 one. In live creatures, lauric and myristic acid are elongated into palmitic acid, although myristic acid usually makes a minor contribution to triglyceride fatty acid compositions. Free oleic acid (MUFA with 18 carbon chains, widely present in both animal and nut/seed/olive/vegetable fats) and linoleic acid (omega-6 PUFA, 18 carbon chains, ubiquitous in vegetable fats but only present in small quantities in animal fats) were both inhibitory, but mostly of the type 1 rather than type 2 isoenzyme. Free palmitic and stearic saturated fatty acids (16 and 18 carbon chains, the usual free fatty acids along with oleic acid in live creatures) were not inhibitory at all.

In cultured human benign prostatic hyperplasia stroma, lauric and myristic acids (but not palmitic or oleic acids) were observed to inhibit 5-alpha reductase activity. In cultured human prostate cancer cells (obviously malignant), unsaturated fatty acids generally were observed to inhibit 5-alpha reductase activity, with linoleic acid doing it more so than MUFAs, and arachidonic acid (omega-6 PUFA, 20 carbon chains) and alpha-linolenic acid (omega-3 PUFA, 18 carbon chains) and docosahexaenoic acid (omega-3, 22 carbon chains) doing it even more so, and gamma-linolenic acid (omega-6, 18 carbon chains, rarer than linoleic acid) doing it most of all.

Another "in vitro" experiment found that free fatty palmitic, oleic and linoleic acids all inhibited the binding of testosterone and DHT by albumin, and the binding of testosterone (but not DHT) by SHBG. In an "ex vivo" experiment, using plasma from men and pregnant women that had been heated (the plasma, not the men and women) to destroy binding by SHBG, saturated fatty acids with more than 16 carbon chains did not affect non-SHBG binding of testosterone, but oleic, linoleic and alpha-linolenic acids increased it.

Sperm cells are far more interesting than benign or malignant prostate cells. Docosahexaenoic acid (DHA) and palmitic acid are the most abundant fatty acids in spermatozoa phospholipids, and are present in approximately equal concentrations in those of normozoospermic men. In asthenozoospermic men, there is much more palmitic acid and a much higher ratio of saturated to polyunsaturated fatty acids (gasp).

More research points to a strong link between depleted concentrations of certain PUFAs in sperm and poor functioning of that sperm. Spermatozoa from asthenozoospermic men had more stearic acid, more total saturated fatty acids (gasp) and a higher ratio of omega-6 to omega-3 PUFAs (double gasp). Those from oligozoospermic men had more stearic and total saturated fatty acids and more oleic acid and total MUFAs. Those from oligoasthenozoospermic men had more oleic acid (tsk, tsk). All the abnormal spermatozoa had less DHA and fewer PUFAs, and high ratios of omega-6 to omega-3 fatty acids were negatively correlated with sperm motility, sperm morphology and sperm concentration.

Does the above mean that everyone should guzzle gallons of flax oil (rich in alpha-linolenic acid, ALA) to "improve" their fatty acid compositions? A couple of things that need to be borne in mind are that: (1) although linoleic and alpha-linolenic acid are both minimally converted into the truly essential polyunsaturated fatty acids (arachidonic and docosahexaenoic acids, ARA and DHA) in humans, high intakes of either can deplete (and totally inhibit production of) the longer-chained PUFA from the opposing class (omega-3 or omega-6) thanks to enzymatic competition; (2) the body maintains its desired balance of saturated and unsaturated fatty acids by desaturating some palmitic and stearic acid into MUFAs, but excessive amounts of "de novo lipogenesis" (DNL, induced by high-carbohydrate and low-fat diets, "healthy" medium-chain fatty acids, "healthy" soluble fibre or regular high ethanol intake) can slow down the fatty acid production line and result in higher than normal concentrations of palmitic acid. Although DHA appears to be more important for sperm, arachidonic acid appears to be more important for brain development, and infant formulas containing high quantities of "healthy" medium-chain triglycerides (primarily 8-chained octanoic acid and 10-chained decanoic acid) have led to depleted bodily concentrations of arachidonic acid (or DHA) in the unfortunate infants! When the above factors (e.g. high-carb and low-fat diets or ethanol intake) do not lead to excessive DNL, they are able to compensate by also increasing the activity of enzymes elongating linoleic and alpha-linolenic acid into ARA and EPA/DHA.

High ratios of palmitic acid to DHA in abnormal sperm could be an effect of oxidative damage depleting the DHA, not the cause of the abnormality of the sperm in the first place. Saturated fatty acids are much more resistant to oxidative damage than PUFAs. Indeed, when subjected to spontaneous lipid peroxidation, human spermatozoa that had previously taken up fatty acids lost far more DHA and ARA than saturated fatty acids.

When added to the extracted semen of normozoospermic men, fatty acids in free form (rather than contained in phospholipids) can immobilize sperm in a manner analogous to the action of autoantibodies against sperm cells, and they can do so at concentrations at or below the normal free fatty acid levels of human plasma. However, two of the most abundant free fatty acids, palmitic and stearic acid, had minimal toxicity even at concentrations massively exceeding the normal upper range! Oleic acid (also an abundant free fatty acid) was less toxic than linoleic acid (which can be a fairly abundant free fatty acid when dietary intake is high), and the most toxic of all the fatty acids tested was alpha-linolenic acid (throw away that flax oil)!

In suspensions of human spermatozoa exposed to a vast range of arachidonic acid (ARA) doses, lipid peroxidation, generation of reactive oxygen species (ROS) and impaired sperm motility were observed, while sperm viability remained unaffected. Other fatty acids were tested for ROS generation and (surprise, surprise) palmitic and stearic acid had no effect whatsoever but linoleic acid and DHA did generate ROS and (at the highest doses, particularly of DHA) even managed to impair sperm viability!

ARA and DHA both perform some vital functions. If a diet does not provide most of its fatty acids in the form of vulnerable PUFAs, it's quite possible that adding a little more of both ARA and DHA will have a beneficial effect. However, if PUFAs (of any type and with any ratio of omega-3s to omega-6s) are dominant in the diet in a readily absorbable form, it is likely that an excess of oxidative damage will override any benefits of extra ARA and DHA. The same will be true if an individual is subjected to oxidative stress from another lifestyle or environmental factor. Hyperglycemia is a potent promoter of oxidative stress. If PUFAs dominate the fatty acid composition of a diet that is otherwise low in fat but which does not result in a chronic excess of "de novo lipogenesis" or in glucose and insulin abnormalities from the high carbohydrate intake, it is likely that oxidative damage will not be excessive and that the modest, sporadic bouts of "de novo lipogenesis" will maintain fairly normal amounts of palmitic acid in bodily cells.

High concentrations of PUFAs from extracted oils (e.g. sunflower, safflower) are likely to be far more hazardous than ones contained within the protective case of a nut or seed (or ones accompanied by saturated fatty acids, MUFAs and fat-soluble vitamins in fish and eggs). Consuming meals based on almonds (moderately high in PUFAs) or walnuts (very high in PUFAs) has reduced the susceptibility of plasma to lipid peroxidation 90 minutes later. Pistachio nuts have a similar fatty acid composition to almonds (high-MUFA with a significant minority component of polyunsaturated linoleic acid), and four weeks of generous pistachio nut consumption (63-126g per day) has reduced levels of oxidized LDL compared to a "western" diet and compared to a low-fat modification of the "western" diet (32-63g of pistachios per day only lowered oxidized LDL compared to the low-fat "improved" diet). Compared to baseline, and compared to when taking a 50g rice powder alternative, post-menopausal women taking 50g of sesame seed powder per day (25g of fat, 11g of linoleic acid, 12mg of gamma-tocopherol) had higher gamma-tocopherol levels (not surprisingly) and reduced amounts of thiobarbituric acid reactive substances (TBARS, oxidized derivatives of linoleic acid) in oxidized LDL. As a matter of interest, they also had higher levels of SHBG, lower levels of DHEA-S and higher urinary levels of 2-hydroxyestrone.

Negative Energy Balance, Fasting and Starvation
Although testosterone levels tend to lower when the diet does not contain plenty of readily available energy (e.g. fatty acids and sugars or easily absorbed starches), they also tend to lower in the period immediately after eating, when the body is busy digesting the food. All else being equal, testosterone levels tend to peak early in the morning (after an overnight fast), and large but transient spikes occur in response to physical training (when previously ingested energy is being metabolized). It seems that, if energy intake and expenditure are fairly even overall, testosterone is produced in greater quantities when energy is being broken down (utilized) rather than stored (made available).

What about negative energy balance? It appears that a sustained modest caloric deficit of 15% has little effect on testosterone levels. Previously sedentary identical twins who performed cycle ergometer exercise (50-55% of VO2 max) for 13 weeks while losing a little under a pound per week had higher levels of testosterone (and DHEA-S) afterwards.

Obese men and women (who have lots of stored fat to fall back on) have both presented higher testosterone levels after prolonged adherence to very-low-calorie diets (we're talking half of my breakfast over an entire day here)! However, if the subjects don't have much stored fat to fall back on, caloric deficits sufficient to sacrifice lean tissue cause testosterone levels to plummet. The crucial factor for maintaining testosterone levels during negative energy balance is whether energy is provided by free fatty acids from broken down triglycerides (as in obese subjects) or by glucose from broken down skeletal muscle protein (as in lean subjects performing endurance exercise).

Testosterone was slashed by more than half when negative energy balance and "endurance training" were combined to an extent that produced a loss of fat-free mass. Combining significant caloric restriction with deadlifts, front squats and negative 1-armed pull-ups would doubtless have done the same thing. Severely restricting caloric intake or fasting altogether prior to a wrestling match (or prior to any other sport for which a weight limit has to be met) causes a decline of that which may help to win the contest. Doug Young managed to win the 1977 World Powerlifting Championships after fasting for an entire week, although he did suffer some cracked ribs in the process:


Two weeks of starvation by subjects deemed to be healthy resulted in lowered testosterone and elevated levels of cortisol and catecholamines (noradrenaline/norepinephrine and adrenaline/epinephrine). The testosterone and catecholamine alterations remained after two weeks of recovery.

In young men, three weeks of exposure to high altitude while maintaining caloric balance has led to elevated free testosterone and cortisol. High altitude forces the body to utilize oxygen more efficiently through the stress of hypoxia, a stress comparable to that induced acutely by vigorous exercise, which also co-elevates free testosterone and cortisol. Free testosterone rose after a few days in a group exposed to high altitude while consuming only 40% of their caloric requirements, but this soon dipped along with a cortisol spike. A group given 40% of caloric requirements without exposure to high altitude displayed lower free testosterone levels.

When 34 US Marine Corps officer candidates aged in their mid-20s were fed just under 40% of the calories that they expended during eight days of military exercises, the IGF-1 and testosterone reductions (around 60% for the free versions) were much the same in those fed a low-protein diet (0.9g per kg, actually higher than the level found in many military ready-to-eat meals) as in those fed an ultra-low-protein diet (0.5g per kg, a measly 38g). Soy contributed 42% of the protein given to both groups. Other androgenic hormones plummetted too, including DHEA, but the sulfate version of DHEA (DHEA-S) increased a little in the first four days.

Negative energy balance, vigorous exercise and hypoxia are stresses. The body can produce testosterone to compensate for stress, and this helps to maintain lean tissue for the duration of the stress, but excessive exposure to stresses (especially combined ones) breaks resistance, induces outright catabolism and causes testosterone to plummet and cortisol to soar.

Dietary Protein and Supplemental Amino Acids, BCAAs, Creatine etc.
As mentioned above, testosterone secretion is a means by which the body can protect itself against the catabolism of skeletal muscle amino acids induced by vigorous exercise, negative energy balance and other stresses. It is not surprising to learn, therefore, that training-related supplementation with protein, amino acids, carbohydrates (which discourage the body from catabolizing amino acids for gluconeogenesis) and creatine (which provides a source of anaerobic energy for the skeletal muscles) has reduced or prevented the catabolism of skeletal muscle amino acids and its associated stress response. 

The effect that the above has on testosterone levels depends on whether the stress being guarded against is moderate and/or singular or intense and/or a synergism of many stresses. Since testosterone production is part of the adrenal stress response that normally lasts for just a short time following challenging endeavours, supplementation around training sessions often results in a small diminishment of testosterone levels despite also enhancing some measures of performance. If an intervention improves performance at an endeavour for which testosterone is helpful, an accompanying mild drop in testosterone should not be a concern. On the other hand, more extreme stresses deplete testosterone levels (probably due to inhibiting reproductive functions), and supplementation under these circumstances is liable to return testosterone levels closer to normal.

Pre- or post-workout protein or amino acid intake has blunted transient free testosterone (and cortisol) elevations but has often resulted in long-term improvements in measures of performance. The same is true for post-workout carbohydrate consumption. Both carbs and low-fat chocolate milk reduced testosterone responses to training in this example, but both (especially the milk) were linked with increases in "fat-free soft tissue".

The branched-chain amino acids (BCAAs, valine, leucine and isoleucine, which together contribute a large portion of skeletal muscle amino acids) spare skeletal muscle tissue from oxidation during vigorous exercise, and supplementation with 50mg per kg of bodyweight per day of leucine (the BCAA that most spares skeletal muscle) had no effect on testosterone levels (which were significantly elevated in the fasting state in male track and field power athletes and sprinters training with sprint intervals who were consuming a mere 1.26g of protein per kg of bodyweight).

Creatine, another substance that often leads to improvements in measures of performance, has elevated resting testosterone (from the free fraction, since neither SHBG nor cortisol were altered) in collegiate American football players over 10 weeks of resistance training. On another occasion, creatine failed to prevent testosterone reductions in response to four weeks of designed "overreaching" but was linked with maintained explosive power and maintained squat and bench press strength during the four weeks, and with increased explosive power in the bench press after the two-week tapering period. By contrast, extra amino acids successfully maintained testosterone levels in another group of resistance-trained men undergoing two two-week phases of designed "overreaching", but in this case their improvements in strength were matched by the testosterone-depleted placebo group.

As for testosterone responses to general high dietary intakes of protein, the problem with research on this matter is that it consists either of vague recollections of eating habits or of controlled research that uses ridiculously high or ridiculously low amounts of protein. No one seems to have thought fit to gauge the effect of, say, 20-25% protein in the diet.

A diet providing 44% of calories from protein has lowered testosterone and (to a comparable degree) SHBG, and raised cortisol and CBG in addition to reducing the 5-alpha reduction of testosterone (i.e. lowering concentrations of DHT). Interestingly, very high protein intakes appear, like fibre, to specifically reduce SHBG (which tends to be lowered by things associated with insulin production and/or gluconeogenesis).

Strength athletes reporting a high protein intake had lower basal testosterone levels and lower free testosterone responses to heavy resistance exercise; the opposite association was found for intake of fat. Another food-recollection study among weight trainers found a similar thing, with the PUFAs:saturates ratio and the protein:carbs ratio being negatively linked with testosterone levels. However, protein intakes ranged from 14-33% of calories, fat from 10-32% and carbs from 48-69%, which doesn't tell us much about truly high fat intakes or about the cut-off point for testosterone reductions from a high protein intake, or about the effect of eating, say, about a third of calories from each of the macronutrients.

Taking all of the above into account, I suspect that (1) pre-/peri-/post-training protein/BCAAs/creatine/milk/carbohydrate etc. intakes diminish acute testosterone responses to manageable challenges because they "spare" testosterone from having to fulfil an anti-stress/catabolism function, but that (2) they minimize testosterone depletions in response to overwhelming challenges; that (3) particularly high basal testosterone levels and/or unusually high or prolonged post-training testosterone spikes "spare" the trainer from needing to ingest supplements immediately afterwards to see comparable improvements in performance; that (4) there is a certain high level of basal dietary protein intake at which testosterone production is mildly reduced due to a "sparing" effect; that (5) there is a higher level of basal dietary protein intake at which testosterone production dips significantly as a result of fatigue from the lack of readily available energy (i.e. fatty acids and sugars), as seen on diets high in "complex carbohydrates".

Alcohol (Ethanol), Testosterone Levels and Testicular Functions
Infants and fetuses do not possess the enzymes necessary for processing ethanol, and maternal intake passes through the placenta and hinders the baby's ability to produce testosterone and estradiol. Ethanol in any appreciable quantity is definitely terrible news for fetuses, and other drug-like items of habitual consumption may not be particularly great news for fetuses either (see the section on methylxanthines). Goodness knows what happened in medieval times, when water was taken in the form of weak ale. Actually, we know what happened: they developed the theological concept that some sort of physiological suffering prior to death (such as the dysfunctioning of the liver, heart and other internal organs as encouraged by ethanol overload) was necessary to cleanse a sin-stained soul and reconcile it with God prior to material extinction.

It is actually contested whether any amount of alcohol is harmful during pregnancy (read the Wikipedia entry on "Fetal Alcohol Syndrome" if you're curious), and it is reported that drinking as much as one drink per day (i.e. the amount beloved by red-wine-sipping yuppies) has not been associated with an increased risk of fetal alcohol syndrome. This may be so, but a more important thing to note is that consuming more than 4.5 drinks per week (which translates in Danish terms into more than 54g or 6.75 "units" per week) has been associated (compared to an intake of less than one drink per week) with lower sperm concentrations in male offspring. On the other hand, while the abstract states that "sperm concentration decreased with increasing alcohol exposure", the figures actually show that those exposed to 1-1.5 drinks per week (i.e. 12-18g or c. 2 "units" per week, or c. 2g or 0.25 "units" per day) had non-significantly higher sperm concentrations than those exposed to less than one drink per week, and significantly higher semen volumes and sperm counts than all the other groups.

As for adults, there are "in vitro", animal and epidemiological studies with some very worrying implications for the sexual organs, and others that give a more ambivalent perspective, but the experimental research with adult humans provides a reasonably clean split (so far as hormone levels are concerned) between testosterone-raising just-enough-to-get-pleasantly-drunk quantities and testicle-intoxicating mass binges.

Ethanol, commonly known as "alcohol", is just one of many alcohols that occur in nature; in fact, anything ending in "-ol" (and much else besides) can be considered an alcohol, including cholesterol, ergosterol, vitamins D2 and D3, retinol (the alcohol form of vitamin A), tocopherols and tocotrienols (forms of vitamin E), polyphenols, and sex hormones including androstenediol and estradiol. Ethanol is produced in overripe fruits via fermentation of their sugars, and very small quantities (i.e. similar to the estimated average daily fetal exposure associated with higher semen volumes and sperm counts mentioned above) are produced naturally by the human gut.

Ethanol can be used as a substrate for "de novo lipogenesis" (DNL) in the liver, but most of it is metabolized therein (and in other parts of the body) into acetaldehyde (the substance to which most of the adverse effects of ethanol consumption are attributed) via an enzyme class (alcohol dehydrogenases, ADHs) whose members (particularly ADH1) are also primarily responsible for metabolizing retinol (the storage and transport form of vitamin A) into retinal (also known as retinaldehyde, the aldehyde form of vitamin A). There are also similar, shorter-chained enzymes known by the name "retinol dehydrogenases" (RDHs), and these also play a role in producing retinal(dehyde) from retinol, but (interestingly) they have as much or more to do with metabolizing steroids. Thus, ethanol and retinol are so similar that they share some of the same enzymes, but the amounts of naturally produced ethanol, while small, are actually much larger than the amounts of retinol found in polar bear's liver (which are believed to be toxic). Retinol is stored in the liver in the form of "retinyl esters", a process that is upregulated by cellular retinol-binding protein 1 (CRBP1), which preserves retinol stores under conditions of vitamin A deficiency or over-expression of ADH1 (as is likely to occur when ethanol is ingested).

Ethanol-derived acetaldehyde, regarded as a carcinogen, is subsequently converted into non-carcinogenic acetic acid via aldehyde dehydrogenases (ALDHs), and the same or similar enzymes (particularly ALDH1A1, also known as retinaldehyde dehydrogenase 1, RALDH1) are also responsible for metabolizing retinal(dehyde) into retinoic acid. ADH3 has a role in metabolizing retinal(dehyde) into retinoic acid, and a smaller role in metabolizing retinol into retinal(dehyde), and is abundantly expressed throughout the body. It is important for survival under conditions of vitamin A deficiency, as is ADH4. Retinal, the aldehyde form of retinol, is regarded as non-toxic whereas retinol can exert a toxicity (at a level that can vary greatly depending on interactions other nutrients, especially other fat-soluble vitamins and zinc) that can be enhanced by retinoic acid and "polar retinoid metabolites" (not named after polar bears), both of which are increased (via the liver enzyme P450 CYP2E1) by ethanol intake. Retinoic acid is the oxidized form of vitamin A, is active and is additionally vital as a hormonal ligand for the retinoid X receptor and the retinoic acid receptor (nuclear receptors which are closely related to steroid receptors and which are over-expressed in the brains of ethanol-overfed animals), but cannot be retro-converted to retinol or retinal (which are interconvertible with each other) and cannot promote eyesight or spermatogenesis as does retinal. Ethanol intake increases production (in various locations) of retinoic acid, polar retinoid metabolites and (in mouse embryos) cellular retinoic acid binding protein 1 (CRABP1). Acetic acid, the end product of ethanol metabolism, is found in vinegar and is produced in a process that also involves the production of free fatty acids (and ketone bodies) via the lipolysis of peripheral triglycerides.

In "in vitro" mouse spermatozoa, ethanol concentrations comparable to those in human drinkers have inhibited capacitation and reduced fertilizing capacity - an effect that was attributed to ethanol itself and not acetaldehyde. Rat studies have led to the suggestion that ethanol's testicular toxicity is attributable to the metabolism of ketone bodies rather than to acetaldehyde formation. By contrast, experiments carried out on cultured canine testes (dog's bollocks) have implicated acetaldehyde as an inhibitor of testicular testosterone production. Human chorionic gonadotropin (hCG) was able to stimulate testosterone production in the experimental bollock (which was given both a higher and a lower amount of ethanol) and in the control bollock, whereas acetaldehyde in a dose equal to the lower amount of ethanol inhibited testosterone production in the experimental bollock.

A single-subject case study reports that ethanol metabolism speeds up in response to declining free testosterone levels, and other research indicates that both testosterone and DHT can inhibit the hepatic (liver) conversion of ethanol into acetaldehyde. High testosterone levels have been associated with a liking for ethanol and other exotic substances, and some liver-intact chronic alcoholics enrolled on a 40-day abstinence programme had free testosterone levels equal to control subjects at the start and higher by the end (they were higher at the start and end in those who were dependent on both alcohol and tobacco). In spite of this, the testicles cannot hold out for ever (high serum testosterone levels sometimes disguise low semen testosterone levels, as mentioned in the section including tobacco smoke) and chronic alcoholics typically present abnormalities including specific deficiencies of free testosterone in the testicles (sub-clinical hypogonadism), reduced sperm counts and motilities and volumes, comparable testosterone levels but reduced sexual activity, or outright lower testosterone levels (which can to some extent recover during withdrawal among those free from signs of severe liver disease). Testosterone levels can recover during withdrawal, given an intact liver, but chronic alcoholism is excellent for damaging the liver beyond repair.

Consumption of carbohydrates in excess of the body's demands can also damage the liver, as reflected in higher levels of various liver enzymes. Whereas high carbohydrate intakes usually result in higher triglyceride levels (linked with the liver enzymes and "de novo lipogenesis" that can also follow ethanol intake) and lower HDL levels, steady ethanol intake often increases levels of both and alcoholism massively elevates SHBG, the latter of which is normally depleted along with elevated triglycerides in "metabolic syndrome" conditions such as diabetes. As a matter of interest, squirrel monkeys fed ethanol at 24% of calories in isocaloric replacement of carbohydrates had higher HDL-3 levels and showed no changes in liver enzymes that were considered to be adverse (although I wouldn't recommend doing the same).

What about regular consumption of very small amounts of ethanol, or intermittent consumption of "moderate" amounts of it, or even regular consumption of large amounts in conjunction with generous quantities of nutrients that will protect against its testicular, liver and other toxicity? As mentioned in the vitamin A section that follows, rats maintained testicular functions when being supplemented with vitamin A in addition to an amount of ethanol that damaged the testicles when given alone. Under the right (or wrong) circumstances, it is possible for vitamin A and ethanol to be synergistically toxic, as seen in the livers of abstinent alcoholics who had previously averaged about 200g of ethanol per day (also mentioned in the vitamin A section). On the other hand, co-supplementation with vitamin A and zinc (an important component of dehydrogenase enzymes) did reverse the hypogonadism of another set of abstinent alcoholics. While ethanol is regarded as a toxin and vitamin A as an essential nutrient, it is worth remembering that ethanol intake is measured in grams and vitamin A intake in micrograms (millionths of a gram), and that an amount of vitamin A equalling the amount of ethanol produced endogenously each day in the gut (c. 3,000,000mcg) would be extremely toxic.

The things discussed above, below and in the vitamin A section illustrate how complicated the effect of ethanol is. Some patterns of usage are very unlikely to negatively affect the liver and testicles (despite transiently inhibiting the hypothalamic-pituitary-gonadal axis) and have the potential (certainly over the short-term but not necessarily the long-term) to modestly elevate testosterone concentrations, but other patterns have the potential to do irreversible damage to the liver and testicles while also promoting various cancers.

Experimental research on live humans has found that an acute dose of 0.5g of ethanol per kg of bodyweight (approximately equivalent to a 70kg person drinking a 500ml bottle of Weston's Vintage Special Reserve Cider providing 4.1 units, eight grams equalling one "unit") produces a moderate transient increase in testosterone and its ratio to androstenedione in men (attributed to an alteration of the liver's redox state, with more NADH and less NAD). It has also found that a fractionally smaller dose does the same thing in women (the effect in both sexes lasted until ethanol had been eliminated but was abolished when ethanol elimination was artificially prolonged). 0.7g or 0.8g of ethanol per kg of bodyweight for women and men increases DHEA and pregnenolone (associated with liking of alcohol in men) and decreases progesterone and allopregnanolone (dropping levels of the last associated with liking alcohol and wanting more in both sexes) and raises testosterone in women (but not in men). It should be remembered that things that raise testosterone in women often lower estradiol, and that regular high ethanol intake is sometimes associated with disproportionately higher levels of SHBG, which could diminish the potency of any mild testosterone elevations.

Consuming ethanol (0.83g per kg of bodyweight) after circuit-style resistance exercise had little effect on testosterone or stress hormones except for a small rise in cortisol after one to two hours.

Acute doses of 2g of ethanol per kg of bodyweight taken over five hours in the evening have increased the urinary ratio of testosterone to epitestosterone in men and women (and testosterone levels in women, which weren't measured in men), though not enough to produce a false positive test for exogenous anabolic steroids. Normally, doses higher than a gram per kg of bodyweight lower testosterone levels in men (as outlined below), especially when taken day after day. Although the just-enough-to-get-pleasantly-drunk doses used in the experiments cited in the paragraph above raised testosterone (and/or DHEA) levels, a person typically needs to gradually drink more to achieve the same pleasant effect when drinking day after day, a conundrum that could lead to boredom and disillusionment (if the same dose is maintained) or to consistently drinking larger amounts that will lower testosterone levels and negatively affect the sexual organs.

On one occasion, however, healthy subjects given about 10g of ethanol per hour for 26 consecutive hours (256g in total, higher even than the intake of many chronic alcoholics) had higher testosterone levels (with no effect on cortisol) throughout the experiment. I fear, however, that, had the experiment continued over the long term, the subjects' livers would've been damaged beyond repair, with their sexual organs following not far behind.

Now for experiments showing a negative impact of high ethanol intake on testosterone production. A little more than 1g of ethanol per kg of bodyweight (from 43%-ethanol whiskey) slightly lowered testosterone acutely and lowered it more comprehensively when taken every day for a week. Male beer and wine drinkers consuming from just under 1g to just over 2g of ethanol per kg of bodyweight (from their favoured beverage and in their habitual amounts) have exhibited diminished synthesis of DHT (along with enhanced synthesis of estradiol). Acute doses of 1.5g of ethanol per kg of bodyweight taken over three hours lowered testosterone and elevated cortisol in healthy young males; the same one-off dose given to males in another experiment (and assumed to have a hangover period of 14 hours) had no effect on testosterone but left cortisol elevated throughout the 20-hour period of observation. A dose of 1.75g of ethanol per kg of bodyweight, ingested by eight healthy males within three hours and resulting in maximum blood alcohol concentrations of 1.51g per litre at the four-hour mark, elevated cortisol for at least 24 hours and depleted testosterone over the same time period, despite elevated levels of the gonadotropins (luteinizing hormone and follicle-stimulating hormone) during the second half of the period of testosterone depletion.

Ethanol intake has some caveat-laden potential to raise testosterone levels. The only down sides are that: (a) it and/or its metabolic by-products are testicular and liver toxins (or a carcinogen in the case of acetaldehyde); that (b) their effects are addictive and that more is usually needed to repeat the effect; that (c) alcoholism typically depletes testicular free testosterone while massively elevating SHBG; that (d) ethanol and/or the process of eliminating it increases the turnover (or decreases the absorption) of most essential nutrients, including vitamin A, ascorbic acid and the B vitamins, and most minerals. The sulfites that are usually added to alcoholic beverages actively destroy thiamin (vitamin B1) in the gut, and ethanol itself (along with the sugars that are often added to alcoholic beverages) speeds up the metabolism of thiamin. Drinking alcoholic beverages can also waste time that could be better spent on other things. On the other hand, it can also help to bring on sleep or encourage profound reflective insights or the adoption of an irreverent stance towards stupid things that don't deserve to be taken seriously, and, if most other things are in good order, regularly drinking enough to induce sleepiness and/or sporadically drinking enough to induce a pleasant state of euphoria is very unlikely to damage the internal organs and may even microscopically elevate testosterone levels.

Vitamin A (Retinol/Retinal, Retinoic Acid and Carotenoids)
As mentioned at the end of the introductory paragraph, two receptors associated with vitamin A (the retinoid X receptor, RXR, and the retinoic acid receptor, RAR) are closely linked genetically with the vitamin D receptor, the thyroid hormones receptor and the steroid hormones receptors. Retinoids (in the form of retinal) play a vital role in spermatogenesis and other testicular functions, and are bound by Sertoli and germinal cells. T3, the most active thyroid hormone, is able to stimulate spermatogenesis independently of testosterone (and maybe independently of FSH as well). The thyroid hormones regulate the metabolic rate, and hypothyroidism impairs testicular function, but hyperthyroidism can do the same via negative feedback.

As mentioned in the alcohol section above, many of the enzymes responsible for metabolizing retinol into retinal(dehyde) and/or retinoic acid are also responsible for metabolizing ethanol into acetaldehyde and/or acetic acid. Interestingly, various short-chain dehydrogenase and reductase enzymes, including retinol dehydrogenases (of which there are more than one, as well as a cis-retinol/androgen dehydrogenase), are closely related and are able to interact and share substrates.

Testicular production of testosterone is dependent on the presence of retinal, which is the aldehyde form of the alcohol retinol, just as acetaldehyde is the aldehyde form of the alcohol ethanol. Since ethanol and retinol compete for metabolism by many of the same enzymes, excessive ethanol usage can starve the testicles of retinal and thereby inhibit testosterone production. Judging by experiments on rats, ethanol consumption (36% of calories for 50 days in replacement of dextrimaltose) induces "gross testicular atrophy" and depletes the testicles of retinal and glutathione (an endogenously produced anti-oxidant), but co-administration of vitamin A in the form of retinyl acetate (in a dose ten times higher than the amount given on the corresponding diet low in vitamin A, which was considered to provide the minimum amount to cover the rats' normal, non-ethanol-intoxicated needs) prevents all of these adverse changes and reduces testicular production of malonaldehyde (a promotor of lipoperoxidation) even by comparison to a group of rats given just as much vitamin A without ethanol! (Both ethanol-fed groups were compared to a control group receiving the same vitamin A intake with dextrimaltose rather than ethanol.) However, testosterone levels among the rats given high-dose vitamin A with ethanol were three times lower than those among the rats given the same dose without ethanol, this being due to ethanol's inhibition of the hypothalamic-pituitary-testicular axis (it prevents the enzymes cyclo-oxygenase and lipoxygenase from using arachidonic acid as a substrate for production of prostaglandin E2 and leukotrienes, which prevents the latter two from activating release of LHRH). LHRH stands for "luteinizing-hormone-releasing hormone" (let us hope that they never discover a hormone that releases LHRH). Testosterone levels in the high-retinyl ethanol-fed rats were the same as those in the low-retinyl ethanol-free rats, which were higher than those in the low-retinyl ethanol-fed rats.

A second, larger caveat is that the ethanol dose in this experiment (36% of calories) was the same as that which was found to synergistically induce giant mitochondria and other lesions when given to rats alongside high amounts of vitamin A in other experiments (see the review cited in the next paragraph). 36% ethanol increased liver retinol-binding protein yet greatly reduced liver retinol levels and increased levels in the blood, the kidneys and the testes. Similar giant mitochondria and liver cirrhosis were seen among ostensibly abstinent alcoholic human males receiving modestly high vitamin A supplementation (also see next paragraph), along with depressed levels of luteinizing hormone.

Interestingly, disease-free regular moderate drinkers typically have higher than normal serum and adipose tissue levels of retinol (and lower levels of beta-carotene). A danger with regular moderate ethanol consumption is that a low intake of retinol/retinal and/or convertible carotenoids (alpha- and beta-carotene and beta-cryptoxanthin) could deplete the liver of retinol and eventually deplete the rest of the body of retinal and retinoic acid (retinoic acid is the oxidized form of vitamin A, active and a ligand for retinoid-related nuclear receptors, but unable to induce spermatogenesis). By contrast, intermittent doses of ethanol could leave behind higher than normal quantities of dehydrogenase enzymes, which could increase the amount of retinal available for the testicles without being nearly vigorous enough to deplete the liver of retinol. Constant high ethanol concentrations, however, make it mightily awkward to maintain liver retinol stores. Chronic alcoholics with liver disease have whole-body vitamin A deficiency but, unfortunately, high doses of vitamin A from retinoids or carotenoids can actually increase damage to livers that are already being battered by ethanol. Note that all of the experiments cited in this review in support of the co-toxicity of vitamin A and ethanol were conducted on former alcoholics or on baboons and rats concurrently being fed 50% and 36% of calories from ethanol - although a mere 3000mcg of retinol per day, given to otherwise free-living abstinent alcoholics with sexual dysfunction (who'd formerly averaged 200g of ethanol per day), was sufficient to produce giant mitochondria and liver cirrhosis or fibrosis in some of them (contrary to a reversal of alcoholism-induced hypogonadism with combined vitamin A and zinc supplementation in another trial cited in this one).

It is too simplistic to expect problems caused by ethanol to be solved by massive doses of isolated vitamin A, such as have typically resulted in toxic symptoms in alcoholics with impaired gonads/livers or chronically inebriated baboons or rats. Zinc, which interacts with vitamin A in many ways, also tends to be malabsorbed (along with protein) by those with damaged livers. Zinc deficiency results in reduced synthesis of retinol-binding protein (RBP), meaning that retinol has an enhanced capacity for liver toxicity, and in reduced synthesis of alcohol dehydrogenase enzymes (ADHs, of which zinc is a component), meaning that less retinol (already potentially more toxic to the liver) is converted into retinal. By contrast, zinc deficiency is also accompanied by greater expression of enzymes converting retinal (irreversibly) into retinoic acid, meaning that a specific deficiency of retinal (the form of vitamin A vital for eyesight and testicular functions) is liable to ensue even if vitamin A intake is reasonable.

High intakes of vitamin A have been accused of promoting osteoporosis and fractures, but some superb research indicates (in a nutshell) that (1) vitamin A and vitamin D3 (which stimulates bone growth and increases the absorption of the calcium and phosphorus required for bone mineralization, but which has also been accused of causing soft-tissue calcification) work more synergistically than antagonistically; that (2) each increases the need for the other and protects against the toxic effects attributed to the other; that (3) higher rather than lower amounts of both are better (scuppering lung cancer and calcification in mice when combined), and that vitamin K2 (production of which from vitamin K1 increases in response to vitamins A and D3) is a crucial go-between in their interactions and is responsible for activating proteins relevant to their interactions and for directing calcium into teeth and bones rather than soft tissues. Although these fat-soluble vitamins increase the amount of one another that can safely and beneficially be ingested, that does not mean that their intake can safely expand indefinitely. Weight for weight, they're much more toxic than ethanol!

In the absence of an adequate intake of vitamins D3 and K2, and in the presence of increased retinoic acid levels caused by regular ethanol intake, it is possible that people could have fracture-susceptible bones. Some research indicates that chronic alcoholics averaging nearly 180g per day have lower blood levels of calcidiol (precursor to the activated hormone calcitriol) despite similar estimated intakes, and female rats to whom ethanol was administered intragastrically displayed higher quantities of one of the CYP24 enzymes that de-activates calcitriol. Other research indicates that male chronic alcoholics free of liver cirrhosis and with normal free testosterone levels (selected on the basis of consuming over 40g of ethanol per day, but actually averaging almost 100g per day) have reduced bone mineral density at the lumbar spine, whereas physically active male soldiers averaging more than 24g of ethanol per week (24g in one go may not be enough to induce a state of pleasant drunkenness, or so I'm informed ...) have greater bone mineral density at the femur (upper thigh) compared to abstainers (they also had lower total and free testosterone but higher total and free estradiol levels, which is interesting considering that the aromatization of testosterone is significantly responsible for bone growth around the hips and thighs).

An adequate supply of retinal is clearly crucial for testicular function, but that doesn't tell us much about the potential of high retinoid and/or carotenoid doses to increase and/or decrease levels of various androgens. To the best of my knowledge, beta-carotene has been tested just once, at a massive dose of 300,000 mcg (300mg) per day for 30 days, and had no effect on various androgens or even on retinol levels. Among 102 male young teenagers with delayed puberty, a combination of 6000 IU of retinol per week (c. 220mcg per day, a bit less than can expected to be present in 100g of cheese) and 12mg of iron per day promoted growth as effectively (and increased testicular volume very nearly as quickly) as exogenous testosterone.

Three months of supplementation by males with severe acne with a massive daily dose of 13-cis-retinoic acid (700mcg per kg of bodyweight, or 49,000mcg or 49mg for a 70kg person) reduced conversion of testosterone into DHT in skin cells and promoted a hepatic shift from 5-alpha to 5-beta reduction of steroids. Note that retinoic acid, as given here, is not active for testicular functions.

A similarly large daily dose of retinoic acid (all-trans-retinoic acid, 35,000mcg or 35mg per day), given to male psoriatic patients for three months, reduced levels of free/bioactive triiodothyronine (T3), the most active thyroid hormone. Hypothyroidism can be induced by a low intake or low availability of calcium, and, as mentioned earlier, hypothyroidism slows down the metabolism and impairs testicular function. Vitamin A by itself reduces blood calcium levels whereas vitamin D increases them, meaning that an imbalance between the two in favour of vitamin A could reduce blood calcium levels and induce hypothyroidism. Note that an "International Unit" of vitamin A represents twelve times more micrograms than an "International Unit" of vitamin D, and that a good balance between the two would still entail a much higher amount of vitamin A than vitamin D, microgram for microgram.

Psoriasis is characterized by conspicuous marks on the outer layers of skin cells, most of which are keratinocytes. In "in vitro" human keratinocytes, T3 had no effect on proliferation and differentiation, but vitamin D3 opposed proliferation and promoted differentiation whereas retinoids had a mild but dose-dependent effect to the contrary. Compared to control subjects with similar vitamin D3 status but normal thyroid status, hypothyroid patients with mild vitamin D3 deficiency responded poorly to a single, massive 100,000 IU (2.5mg) dose of vitamin D3, and only administration of the thyroid hormones themselves was able to correct vitamin D3 status (it also increased concentrations of DHEA and DHEA-S). The thyroid hormones (T4 and T3) are derived from the amino acid tyrosine and the mineral iodine.

The above research illustrates the importance of synergistic interactions between different essential nutrients that are vital for general health and for sexual function, and the folly of banking on massive doses of one without having any regard for the others. As we shall see in the next section, retinoids are not the only members of this group of essential nutrients that interact with steroid hormones in fascinating ways.

Vitamin D
Androgens and estrogens are widely regarded as being nefarious promotors of cancer growth, especially in prostate and breast cells, largely on the basis of "in vitro" experiments involving cultured cells. Since one of the main functions of sex hormones is to promote growth, it is perhaps not surprising that they should also promote the growth of cancer cells when the two are examined in isolation outside of a living organism. Does this mean that sex hormones will promote cancer in a healthy body that has a robust immune system and is not deficient in any vital nutrients? Cell culture experiments using steroid hormones and fat-soluble vitamins suggest that this notion may be total and utter twaddle.

Metabolites of vitamin D3 (calcidiol or calcitriol) have an anti-proliferative effect on cultured human prostate cancer LNCaP cells, but this effect can be diminished by an enzyme (24-hydroxylase, CYP24) that converts these metabolites into a form that is not active in these cells. Luckily, however, this enzyme can be downregulated by another biochemical entity, which thereby sustains the anti-proliferative effect of vitamin D3 metabolites. What is the name of this entity? It is none other than the villainous end-baddie itself, 5-alpha dihydrotestosterone (DHT)!

Calcitriol and a non-steroidal anti-inflammatory drug (ibuprofen) synergistically inhibited LNCaP cell growth in a DHT-free medium in another set of experiments, but the anti-proliferative effect was enhanced by adding DHT to the culture medium! Ibuprofen was added due to concerns about the potential hypercalcemic effect of calcitriol. One suspects that vitamin K2 (covered in the next section) would have been more appropriate.

Calcitriol also has an anti-proliferative effect in another human prostate cancer cell line (MDA). The effect is blocked by a pure anti-androgen, though only partly at low doses, indicating the effect of calcitriol to be somewhat independent. The MDA line has very-low-affinity androgen receptors, but calcitriol increased androgen receptor messenger RNA (ribonucleic acid)!

The synergism between vitamins A and D3 also appears to be crucial to the relationship between sex hormones and cancer cells, again judging from cell culture experiments. Calcitriol and retinoids (cis-9 and all-trans retinoic acid) both promoted the differentiation of LNCaP cells with or without androgens, but the retinoids (and even calcitriol under certain conditions) promoted the growth of the cells without androgens, whereas both vitamins (especially calcitriol) inhibited the growth of the cells with androgens! In addition to this, calcitriol approximately doubled the number of androgen receptors!

Interestingly, another biochemical entity which enhances the effect of vitamin D3 in prostate cancer cells is "alpha-tocopheryl succinate", a "redox-silent" form of vitamin E that has neither anti-oxidant nor pro-oxidant activity.

Given that DHT downregulates the enzyme (CYP24) responsible for inactivating vitamin D3 metabolites that are able to antagonize cancer progression under "in vitro" conditions, and given that estradiol is blamed for contributing to breast cancer just as DHT is blamed for contributing to prostate cancer, is it possible that a similar relationship exists between vitamin D3 and estradiol? Indeed it is. In human breast cancer cells, 17-beta-estradiol downregulated the activity of CYP24 and also upregulated the activity of an enzyme (CYP27B1, 25-hydroxyvitamin D3 1-alpha-hydroxylase) that produces the most bioactive form of vitamin D (calcitriol)!

Vitamin K
Since vitamin K interacts with vitamins A and D, and since vitamins A and D and their receptors and enzymes interact in complex ways with steroid hormones and their receptors and enzymes, a question that quickly arises is whether there is any indication that vitamin K could interact in some way with the steroid hormones. The answer is yes: both androgen-binding protein (ABP) and SHBG are closely related to the "globular domain" of "S family" proteins, the latter of which are dependent on vitamin K. Specifically, the COOH-terminal domain of Protein S is homologous to ABP and SHBG. What this means is not clear (to me), but it is interesting.

Although vitamin K is typically characterized as being coagulatory, the S family of proteins (one of many families activated by vitamin K) are actually anti-coagulatory, and they are present in the testicular Leydig cells, although they do not bind androgens. Protein C, another protein with an anti-coagulatory effect (it degrades coagulatory factors Va and VIIIa), is also dependent on vitamin K (and on Protein S possessing a single beta-chain, heterozygous deficiency of which promotes thrombo-embolic events).

Interestingly, deficiency of Protein S appears to go hand in hand with testosterone deficiency and venous thromboses. Administration of synthetic ethinyl estradiol (but not natural 17-beta estradiol) to male-to-female transsexuals has been linked with venous thrombosis via depleted total and free plasma Protein S - the opposite of what is seen in female-to-male transsexuals given testosterone.

Vitamin K2 acts as a ligand and, in bone cells, is able to activate the "steroid and xenobiotic receptor" (SXR), a nuclear receptor belonging to the NR1I sub-family along with the vitamin D receptor and the "constitutive androstane receptor" (CAR). The vitamin D receptor (closely related to the pregnane X receptor) is thought to be the original NR1I gene.

The drug warfarin, which is enthusiastically administered to people with excessively coagulatory blood, inhibits the enzyme (vitamin K epoxide reductase) responsible for recycling vitamin K after it has "carboxylated" (activated) proteins. As mentioned above, anti-coagulatory as well as coagulatory proteins are dependent on vitamin K (they cannot be activated without it). Osteocalcin, a protein hormone vital for bone growth and dental health, is also carboxylated by vitamin K, as is the protein matrix Gla, under-carboxylation of which contributes to arterial calcification. Both osteocalcin and matrix Gla appear to be carboxylated much more effectively in response to vitamin K2 rather than vitamin K1.

Since vitamin K2 is more important for carboxylation than vitamin K1, and since warfarin inhibits the enzyme responsible for recycling vitamin K following carboxylation, this implies that the more crucial K2 form will be by far the biggest casualty of warfarin use. This may help to explain why use of warfarin has led to increased plasma levels of vitamin K1, an increase that was accompanied by decreased quantities of folate (vitamin B9) in red blood cells (erythrocytes). Four weeks after discontinuing warfarin use, one group of patients exhibited venous thrombosis along with reduced erythrocyte folate and hyperhomocysteinemia (which the authors politically attributed to a lack of leafy green vegetables rich in vitamin K1 and not to use of warfarin). High homocysteine and low folate plasma levels have been correlated with the calcification of arterial plaque, and high homocysteine and low vitamin B12 levels have been correlated with reduced bone mineral density! Hmmm. Estimated high intakes of vitamin K2 (menaquinone), but not those of vitamin K1 (phylloquinone), have been correlated with reduced coronary heart disease (and reduced all-cause) mortality among nearly 5000 subjects of both sexes, and with reduced incidence of CHD (mortality isn't mentioned in the abstract) among more than 16,000 women! Are you starting to connect the dots? Low intake of leafy green vegetables, my arse!

Vitamin K is vital for healthy bones, primarily in the form of vitamin K2. Under "in vitro" conditions in human primary osteoblast cells, vitamin K2 managed to promote mineralization even when hindered by the presence of warfarin. It also encouraged the conversion of osteoblasts into mature bone cells (osteocytes) and diminished the cells' ability to produce osteoclasts (breakers-down of bone), again with little or no sensitivity to warfarin. All of these things may have been due to gamma-carboxylation and/or activation of the steroid and xenobiotic receptor (SXR).

Vitamin K2 supplementation in live humans has resulted in reduced production and reduced proliferation of liver cell (hepatocellular) carcinomas (HCCs). In "in vitro" experiments, HCC cells overexpressing the steroid and xenobiotic receptor (SXR) were less motile and proliferated less rapidly. Adding vitamin K2 to the medium (and thereby increasing activation of the already overexpressed SXR) made the HCCs even more impotent!

Thus vitamin K2, like vitamin D3 (with which it interacts synergistically), is crucial for bone health and antagonizes the proliferation of certain cancer cells via mechanisms that involve steroid hormones and/or their receptors!

Vitamin E (Tocopherols and Tocotrienols)
Besides vitamins A, D and K, the other fat-soluble vitamin is vitamin E, which occurs in the form of four tocopherols and four tocotrienols (alpha-, beta-, gamma- and delta-), the best studied of which are alpha-tocopherol (the form most often found in supplements) and gamma-tocopherol (the most abundant form in dietary sources). As mentioned at the end of the vitamin D section, a synthetic form of vitamin E, alpha-tocopheryl succinate, which has little or no anti-oxidant (nor pro-oxidant) activity (which does not prevent it from being promoted for its "anti-oxidant" capacity on websites selling nutritional supplements), has synergistically enhanced the effect of vitamin D3 in antagonizing prostate cancer cell proliferation (in cultured prostate cancer cells and in unfortunate mice to whom such cells were xenografted).

Vitamin E and specifically alpha-tocopherol are believed to be essential for sexual function on the basis of experiments from the 1920s which observed infertility (most sterile in the first generation, the rest in the second) among rats that otherwise grew and developed perfectly well on a diet of casein (protein), cornstarch, yeast, lard and butterfat. The sterility was prevented by adding lettuce to the diet or by increasing the proportion of butterfat to 24% of the diet.

In my opinion, the above experiments are not sufficient to show that alpha-tocopherol is essential for sexual functions. According to the USDA nutrient database, various forms of lettuce all contain only small amounts of alpha-tocopherol but massive amounts of vitamin K1. Butterfat (which also corrected the deficiency when fed in large enough amounts) also contains small amounts of alpha-tocopherol as well as small amounts of vitamins K1 and K2. Dietary fat increases the conversion of beta-carotene into vitamin A and could also do the same for vitamin K1 into K2, and vitamins A and D3 have increased the conversion of vitamin K1 into K2 in experiments on mice. Butter is also rich in vitamin A and, as mentioned in the vitamin A section, testicular retinal serves an anti-oxidant function by preserving levels of endogenously produced glutathione. In theory, then, these rat experiments could be interpreted as proving the essentiality of vitamin K and/or of preserving glutathione and anti-oxidant capacity in the sexual organs via larger doses of vitamin A.

In spite of the skeptical view given in the paragraph above, the essentiality of alpha-tocopherol for fertility is said to have been established in experiments spanning from the 1920s to the 1950s, the vast majority of which I have not looked into. Alpha-tocopherol may well be absolutely essential for fertility, and it and other forms of vitamin E are all very useful in amounts that can be obtained from dietary sources. However, mass-dose supplementation with 10,000 international units reduced fertility in female rats.

In other rats, who were deficient in selenium, massive doses of alpha-tocopherol failed to normalize testicular function. Selenium too is an "anti-oxidant", and it is clear that a robust anti-oxidant capacity is essential for healthy sexual functioning. Like most anti-oxidants, selenium is also able to promote oxidative damage, but it is able to recycle the endogenous anti-oxidant glutathione back from oxidized into "reduced" form.

There is one report of an unspecified dose of supplemental vitamin E (form not specified) increasing plasma testosterone in "normal male subjects".

The problem with alpha-tocopherol is that massive-dose supplementation with it has been linked with marginally increased mortality in many human trials, which may well have something to do with it depleting the body of gamma-tocopherol and delta-tocopherol. Controlled feeding experiments on healthy pigs, in contrast to pigs with pre-existing endogenous oxidative stress, have observed impaired endothelial function in those given dl-alpha-tocopherol, despite accompanying reductions (or non-alterations) in traditional markers of lipoperoxidation. Oxidative damage occurs not only to lipids but also to proteins (nitrosation), and there was clear evidence of increased nitrosation in these pigs.

Under "in vitro" conditions, and under "ex vivo" conditions in blood extracted from live humans (including ones with homozygous hereditary deficiency of alpha-tocopherol), alpha-tocopherol acetate antagonized lipoperoxidation under conditions of high oxidative stress, but did so under conditions of low oxidative stress only when accompanied by large amounts of vitamin C (ascorbate). It had a pro-oxidant effect under conditions of low oxidative stress when vitamin C was absent. Alpha-tocopherol can itself be oxidized into a pro-oxidant form, and other anti-oxidants, especially vitamin C, are able to regenerate it into anti-oxidative form. Protein oxidation was not examined.

The processes of lipid and protein oxidation overlap. A mutagenic nitrogen and oxygen reaction product, peroxynitrate, is formed via the activation of phagocytes, a process that strongly encourages their accumulation with oxidized LDL on arterial walls. When alpha- and gamma-tocopherol were compared, gamma-tocopherol was not more effective at inhibiting lipoperoxidation in isolated LDL but was more effective at inhibiting lipoperoxidation in liposomes.

Other research shows that alpha-tocopherol reacts with nitrogen dioxide to form a nitrosating product called tocopheroxide. By contrast, gamma-tocopherol reacts with nitrogen dioxide to form nitric oxide, a gas which promotes endothelial vasodilation and thus widening and relaxation of the arteries and reduced blood pressure. Gamma-tocopherol was also a more potent inhibitor of neoplasia.

Interestingly, both gamma-tocopherol and the anti-oxidant epicatechin have been shown to promote endothelial nitric oxide production, and both are abundantly present in dark chocolate. Nitric oxide has a close relationship with testosterone in the testicular process leading to penile erection, and testosterone itself promotes vasodilation. Dark chocolate and cocoa also contain an abundance of the methylxanthine theobromine. Methylxanthines (as discussed in their own section) can improve sexual function, but massive doses in purified form (higher than can be realistically obtained from natural dietary sources) can have the reverse effect and it is possible that even smaller doses (as obtained from dietary sources) are able to have an adverse effect on fetuses. It is likely that excesses of methylxanthines and exogenous, non-essential "anti-oxidants" are both capable of reversing the beneficial effects of more modest doses.

The things considered above suggest that large supplemental doses of alpha-tocopherol are best avoided, that any of the much more modest amounts of alpha-tocopherol obtained from dietary sources are good, and that any gamma-tocopherol from dietary sources is possibly even better.

Vitamin C (Ascorbic Acid), Glutathione and Uric Acid
I have found very little research directly linking vitamin C (ascorbic acid) and sex hormones. However, it is clear from the things discussed in the sections on vitamins A and E that oxidative damage is a mortal foe of sexual function. Vitamin C is both an essential nutrient and a potent anti-oxidant, and most animals apart from "higher" primates, guinea pigs, some bats and birds and at least one nematode parasite (creatures which have access to high amounts from food) take care to produce massive amounts of vitamin C (ascorbic acid) from glucose (the lucky bastards). Creatures who don't produce their own lack the final enzyme (L-gulono-gamma-lactone oxidase) in the metabolic process leading from glucose to ascorbic acid.

Many creatures, including those that don't produce their own ascorbic acid, are also able to produce the anti-oxidants uric acid and glutathione. Uric acid is derived from purines (including those derived from methylxanthine alkaloids), and blood levels increase in response to a large intake of protein (which provides exogenous purine) or an excessive intake of fructose or other sugars (metabolism of which encourages the endogenous production of purines and the scavenging of phosphate from ATP, which then needs to be regenerated by the release of phosphate from bone). Excess uric acid concentrations are linked with conditions such as gout (which mainstream nutritionists eagerly blame on dietary protein, despite the obvious objection that protein-rich foods provide an exogenous source of purine, not to mention phosphorus, which relaxes the ATP-diminishing effect of fructose). An intake of 500mg per day of vitamin C reduces levels of uric acid, more so in each higher quartile of initial uric acid.

Glutathione is derived from the amino acids glycine, glutamic acid and cysteine. Cysteine, along with another "sulfuric" amino acid (methionine), is also crucial for the health of tendons and other connective tissues. Methionine and cysteine are most abundant in animal sources of protein, are known as sulphur-containing amino acids (SAAs) and increase the urinary excretion of calcium. Fortunately, animal sources of protein also increase the absorption of calcium, and minerals such as phosphorus and potassium (also abundantly present in non-purified sources of animal protein) discourage the urinary excretion of calcium, meaning that the net effect of natural sources of animal protein on calcium balance is either neutral or (if they are rich in calcium itself) positive - again despite the idiotic insistence of mainstream nutritionists to the contrary.

Fertile males (smokers or non-smokers) have higher ascorbic acid levels in seminal plasma than infertile ones, and ascorbic acid levels also correlate with a higher percentage of sperm with normal morphology. Oligozoospermic males have reduced spermatozoa levels of glutathione compared to normozoospermic ones, and higher intracellular glutathione levels correlate with the capacity to penetrate bovine cervical mucus!

Not surprisingly, low levels of vitamin C and glutathione have been linked with low testosterone levels and elevated levels of TBARS indicating oxidative stress and testicular toxicity among male alcohol abusers aged 20-40.

B Vitamins
I have found very little concerning the direct effect of various B vitamins on sex hormone levels and related matters. A few interesting things to note are that (1) male mice lacking a certain thiamin(e) (vitamin B1) transporter are infertile; that (2) immunization against the riboflavin (vitamin B2) carrier protein (RCP) promotes infertility in various species; that (3) niacin (vitamin B3, which can be produced from the amino acid tryptophan) is the source of the NAD+ and NADH which are necessary for the metabolism of substances including steroids; that (4) pantothenate (vitamin B5) supplementation has reversed health defects including immature sperm in male mice fed a B5-deficient diet; that (5) megadoses of pyridoxine (precursor to the active form of vitamin B6, given at doses from 125mg to 1000mg per kg) hypotrophy the testicles and reduce sperm counts in male rats (which doesn't necessarily mean that more normal high doses would do the same) and that pyridoxine increases 5-alpha reductase in cultured rat hypothalamus and pituitary regions whereas pyridoxal (its active derivative) inhibits it, and that vitamin B6 deficiency increases the uptake of steroid hormones into cultured rat uterus slices (possibly via increased receptors or enzymes) but repletion increases it even more; that (6) three months of massive 15mg (15,000mcg) per day supplementation with folinic acid (a synthetic form of folate, vitamin B9) increased sperm counts and motility in male partners of infertile couples (especially in those 17 of 65 whose partner subsequently conceived), and that (7) large doses of synthetic vitamin B12 (methylcobalamin, 1500mcg per day for anything from four to 24 weeks) were judged to have improved the sperm parameters of 11 out of 26 non-azoospermic infertile males (but to have impaired them further in four of the others).

B vitamins are water-soluble (like vitamin C) and can therefore offer some protection against the strong diuretic (water-excreting) effect of certain interesting drug-like items of diet (methylxanthines, ethanol) that have an ambivalent relationship with sex hormones and sexual function.

Macro-Minerals (Sodium, Potassium, Phosphorus, Calcium, Magnesium) and Boron
The two most abundant minerals in the human body are sodium and potassium. I have found no research looking directly at the effect of these on sex hormone levels, but it is important to note that production of the "mineralocorticoid" stress hormone aldosterone (a steroid hormone derived from cholesterol like cortisol and the sex hormones) increases in response to low sodium levels and high potassium levels. Potassium is a component of the glycogen stored as fuel for vigorous muscular exercise, and a high intake along with vigorous exercise is therefore less likely to lead to high blood levels (hyperkalemia) and excess production of aldosterone (more of the potassium being directed into skeletal muscles).

Although sodium and its densest source (salt) are popularly stigmatized as promoting hypertension (raised blood pressure), the balance between sodium and potassium is more important than the absolute amount of sodium ingested, and the body is able to tightly regulate the balance between the two by excreting excesses via the kidneys (something that is easier to do when there are more than adequate amounts of both, since two wrongs do not make a right). Of the two, it is sodium that is the most abundant mineral in the human body. (The things discussed briefly in this paragraph would make a good topic for a future article.)

There are two main pathways to sodium-potassium imbalance. Since salt is very useful for preserving foods, relying primarily on conveniently preserved foods rather than fresh ones (as is liable to happen in "western" societies) can lead to an imbalance on the side of sodium (which can be exacerbated further by the use of table salt consisting of almost pure sodium). Because of the greater familiarity of sodium-slanted imbalance in "western" societies, resultant hypertension is much more stigmatized as a serious health concern than is hypotension resulting from the opposite imbalance.

On the other hand, the over-reliance on fresh foods of a terrestrial/inland origin (rather than a marine/coastal one) can lead to an imbalance on the side of potassium (sodium is most prevalent in the sea). Elephants relying mainly on inland vegetation for food go to the trouble of manoeuvring their vast bodies through tight underground caves at night just to get a lick of a supposedly toxic substance. Despite the negative stigma attached to purified table salt, natural sea salt contains fairly decent quantities of other minerals besides sodium, especially magnesium and iodine (the precursor to the thyroid hormones, as related above and below).

So, although there is little or no direct research concerning the effect of sodium and potassium on sex hormone levels, it is clear enough that an imbalance either way leads to health concerns that are very unlikely to be conductive to optimal physical performance or sexual function, and that a sodium-starved body gravitates towards producing more stress hormones (and thus probably less sex hormones) while potassium is a necessary component of the glycogen that is required for optimal physical performance. As well as utilizing potassium-containing glycogen, exercise (especially of the endurance variety) also results in the loss of bodily sodium (hyponatremia).

Likewise, I have found very little regarding any direct relationship between phosphorus and sex hormone production, but it too has a crucial role in optimal physical performance, one that is taken for granted due to its ubiquity in the food supply. Phosphorus is a component of adenoside tri-phosphate (ATP), adenosine di-phosphate (ADP), adenosine mono-phosphate (AMP) and creatine phosphate or phosphocreatine (CP/PC), all of which are vital for energy production, and it is a component of teeth and bones along with calcium (the latter of which is more emphasized due to the narrower range of foods from which it can be obtained in high, well-absorbed quantities). In other words, it is vital for all the things for which people desire to have an elevated production of sex hormones.

Among post-menopausal women being fed in a metabolic unit, a boron-supplemented diet (3mg per day plus 0.25mg from the diet) reduced urinary excretion of calcium and magnesium and resulted in elevated levels of both testosterone and estradiol - patterns that were more pronounced when the diet was very marginal in magnesium (116mg per 2000 calories versus an additional 200mg from a supplement). On another occasion when it was given (3mg per day) to post-menopausal women in a metabolic unit for three weeks as part of an otherwise identical (and truly awful) diet, those women had consistently positive calcium balances, progressively more positive (or at least less negative) potassium balances, and progressively more negative magnesium balances, along with no interesting alterations to any hormone level. Does the above mean that magnesium restriction is a desirable thing? I doubt it (see the next-but-one paragraph). However, it is possible that boron is an essential nutrient in its own right.

Boron given to healthy males for four weeks (10mg per day) resulted in an ambiguous trend for higher testosterone levels and an unambiguous elevation of estradiol levels. Males in their 20s who took 2.5mg of boron per day while training on four days per week for seven weeks had higher testosterone levels and increased squat 1-rep maxes at the end of the trial, but the same outcomes were seen in the placebo group, indicating that "structured resistance training" is far more important than trendy supplements.

Since marginal magnesium status seems to be linked with marginally more sex hormone production and more positive calcium balance in post-menopausal women following appalling research unit diets, it must be a bad idea to ingest large amounts of magnesium, eh? Think again! Taking 10mg of magnesium per kg of bodyweight per day (i.e. 700mg per day for a 70kg person) resulted in elevated total and free testosterone levels in sportsmen practising taekwondo for 90-120 minutes per day. A similar effect was seen in a similar study using 35mg of calcium per kg of bodyweight. It is possible that magnesium interferes with the binding of testosterone by SHBG.

Trace Minerals (e.g. Iodine), Goitrogens and Toxic Metals and Ions (e.g. Fluoride)
As mentioned at the end of the introductory section and in the section on vitamin A, the thyroid hormones (especially triiodothyronine, T3) are strong promotors of spermatogenesis and their receptors belong to a family of nuclear receptors including those for fat-soluble vitamins and steroid hormones. The thyroid hormones are derived from the amino acid tyrosine and the essential mineral iodine (abundant in sea water, with increasingly diminished amounts in soil further inland). The best dietary source of iodine is natural sea salt, which should also be a good source of magnesium and possibly other minerals in addition to the abundant sodium. Table salt is almost purified sodium and contains no other minerals, although iodine is sometimes added back in.

Cruciferous vegetables (Brassicaceae or Cruciferae) and various others contain entities known as goitrogens (or glucosinolates or isothiocyanates), which prevent the body from utilizing iodine, resulting in hypothyroidism and other symptoms of deficiency. Levels of the very same properties in the urine of a multi-ethnic cohort have been linked with a reduced incidence of colorectal cancer, but a high intake of them by mildly iodine-deficient females has been linked with an exceptionally high rate of thyroid cancer. Excess iodine and hyperthyroidism can cause similar problems, which may explain why intakes of iodine and seafood were positively associated, and those of cruciferous and other goitrogenic vegetables negatively associated, with the incidence of thyroid cancer in Hawaii (a group of small islands on which all locally produced food is bound to be abundant in iodine). These things provide an excellent example of how complicated are the interactions between all the different chemicals encountered in the diet and the environment.

Iron does not appear to have any prominent role related to sex hormone production and the functioning of the sexual organs, but deficiency of it (anemia) is customarily accompanied by impairments of both, as is only to be expected, and higher levels of hemoglobin (the oxygen-transport protein of which iron is a critical component along with folate and vitamin B12) have been correlated with total and free testosterone levels in a group of nearly 500 men aged 30-95. As mentioned in the vitamin A section, a daily 12mg iron supplement along with around 220mcg per day of retinol was very nearly as effective as exogenous testosterone at increasing the testicular volume of young men with delayed puberty. Any amount of testosterone would be pretty useless in the complete absence of an essential nutrient.

As mentioned in the vitamin E section, selenium is very important for testicular anti-oxidant capacity (and thus proper sexual functioning). Selenium is highly effective at recycling endogenous glutathione from oxidized (pro-oxidant) back into "reduced" (anti-oxidant) form. Taking 200mcg per day of selenium (along with 600mg of N-acetyl-cysteine) for half a year increased testosterone levels and improved the semem parameters of infertile men. Eating just two Brazil nuts is sufficient to significantly improve selenium status.

Copper is thought to be crucial for reproductive functions, even if I haven't managed to find much relevant research on live humans. If you do manage to raise testosterone naturally, it is worth bearing in mind that testosterone may intensify copper deficiency. As a matter of interest, fructose seems to increase copper absorption but may also hinder its utilization once in the body. Dark chocolate is rich in copper, iron, zinc and magnesium (and manganese), although it's likely that their absorbability is very limited.

Manganese is thought to be toxic to the testicles, though I've found no relevant research on live humans. In cultured rat Leydig cells, testosterone production stimulated by human chorionic gonadotropin (hCG) was dose- and time-dependently decreased (along with "steroidogenic acute regulatory protein", or StAR, and related enzymatic processes) by manganese. However, very small amounts of manganese appear to protect sperm extracted from human males by increasing levels of reduced and lowering levels of oxidized glutathione.

Among Croatian males aged 20-43, high blood levels of the toxic metal cadmium were associated with higher levels of testosterone (and estradiol) but also with decreased sperm motility. 2.5mg of cadmium per kg of bodyweight per day reduced the testicular mass of some unfortunate rats, although the effect was reversed by both zinc and diallyl sulfide from garlic (also found in other members of the Allium genus, which includes onion, shallots, chives and leeks). Among Croation males aged 19-52, a similar pattern (elevated testosterone and estradiol but abnormal sperm and depleted seminal plasma zinc) was seen in conjunction with high blood lead levels.

Diallyl sulfide from Allium plants (which can increase testosterone production via luteinizing hormone, as discussed in the relevant section) is (as the name suggests) a sulfurous entity, and zinc (vital for healthy sperm) often occurs in sulfurous form, but exposure to sulfur mustard produces sperm abnormalities without necessarily causing testosterone levels to drop. If testosterone levels do initially drop, the sperm abnormalities remain long after testosterone has returned to normal. Sulfites/sulphites used to preserve certain foods and alcoholic beverages are able to destroy thiamin (thiamine, vitamin B1), and ethanol itself antagonizes its intestinal absorption (but may reduce its urinary excretion despite being a diuretic and despite the vitamin's water-solubility).

One of the most controversial minerals-cum-toxins is fluoride, which is touted as being good for dental health and which is consequently added to most toothpastes and (in some parts of the world) to the water supply. I have not examined in any detail the relationship between dental health and different forms of fluoride (like all minerals, it occurs in many different complexes, some of which may have different effects from others), but of what consequence are the teeth when the testicles themselves are in mortal peril? It is quite clear that fluoride is a testicular toxin.

Men who had lived for more than five years in a district with nearly 4mg per litre of fluoride in the drinking water had (in comparison to those inhabiting a region with less than 1mg per litre) higher luteinizing hormone levels but lower testosterone levels. The women had higher testosterone but lower estradiol levels. Men drinking the same water as skeletal fluorosis patients, without exhibiting skeletal fluorosis themselves, had testosterone levels intermediate between the patients and control subjects whose drinking water contained less than 1mg per litre. Men exposed to 3-27mg of fluoride per day had, compared to those exposed to 2-13mg per day, higher follicle-stimulating hormone (FSH) and lower free testosterone levels.

There is one report that fluoride exposure up to 250 parts per million in drinking water fails to effect reproductive functions in male and female rats or their offspring (given the same treatment), but another experiment found decreased spermatogenesis and steroidogenesis in neonatal male rats whose mothers had been given fluoride at a measly 4.5 or 9 parts per million in the drinking water.

Lipid peroxidation in the testicles, livers and kidneys was seen in male rats who had been given 20mg of fluoride per kg of bodyweight per day. An identical dose of vitamin E protected against the effect, as did a much smaller alternate-day intraperitoneal injection of testosterone. More male rats, given the same dose, had lower testosterone levels, diminished activity of steroidogenic enzymes, smaller testicles, reduced sperm counts and less mature sperm in conjunction with elevated markers of oxidative stress.

What this indicates for the potential hazard of brushing one's teeth with fluoridated toothpaste (most of which gets spat out) is not clear.

Zinc
Zinc is crucial for steroidogenesis and spermatogenesis. Low serum and seminal plasma zinc levels correlate with low testosterone levels in oligospermic and azoospermic infertile males. Experimental human zinc deficiency results in lowered testosterone levels, poor sperm function, weakening of the immune system and a reduction of lean body mass. Concentrations within cells are depleted before concentrations in plasma, meaning that deficiency could go undetected. Zinc supplementation typically increases testosterone levels in infertile, ill, healthy and vigorously exercising male subjects of all ages.

It increased the previously depleted testosterone levels of males with sickle cell amaemia. In addition to testosterone and DHT levels, it increased sperm counts in some males with idiopathic infertility, and some of the wives were subsequently fertilized, although there was no fertility among those who had presented not-so-low testosterone levels to begin with. Since serious zinc deficiency causing infertility typically causes testosterone levels to plummet, it is likely that the infertility of the males with not-so-low testosterone levels was due to something other than dietary zinc deficiency.

Cellular zinc concentrations correlated with serum testosterone concentrations among a sample of 80 American men aged 20-80. Dietary zinc restriction reduced testosterone levels in four young men, and zinc supplementation increased testosterone levels in nine elderly men who were initially somewhat deficient in zinc.

11 men with an average age in their late 20s were monitored on low-zinc diets (1.44mg per day) from a mixture of regular and semi-synthetic foods under metabolic ward conditions. The diets were given in five five-week phases (separated by four-week washout phases) and differed in the amount of (probably very bioavailable) supplemental zinc that was added to them, which was a low amount (1.4-4.4mg) except for in the final, non-random phase (10.4mg, the same amount as fed in the washout phases). Compared to the 10.4mg dose, testosterone levels and semen volumes were reduced on the lowest dose, and seminal zinc levels remained intact only because of reduced levels of zinc per ejaculate on the 1.4mg, 2.5mg and 3.4mg doses.

A zinc supplement called ZMA did not increase urinary testosterone concentrations when given to regularly exercising men who were judged to be consuming a little more or a little less than the so-called "recommended daily amount" of zinc (usually 10-15mg). On the other hand, four weeks of a massive daily oral zinc dose of 3mg per kg of bodyweight (210mg for a 70kg person) increased total and free testosterone levels in male wrestlers aged about 18, and these increases were present both at rest and following exhaustive exercise.

Drink up thy zinc!

Garlic, Onion and Allium Plants Containing Diallyl Sulfide
Both zinc and diallyl sulfide prevented the reductions of testosterone levels and testicular weights seen in adult male rats given 2.5mg of cadmium per kg of bodyweight. Diallyl sulfide is found in plants of the Allium genus, which includes garlic, onions, shallots, leeks and chives.

In another experiment on rats, adding garlic (Allium sativa) powder at 0.8% to a diet with a high protein intake (25% or 40% casein) increased testosterone and lowered levels of corticosterone (the main corticosteroid hormone in most vertebrates other than humans, in whom it is an intermediate between pregnenolone and aldosterone). Adding it to a low-protein diet (10% casein) had no effect. In a follow-on experiment, diallyl disulfide was observed to dose-dependently increase levels of luteinizing hormone.

Juice extracted from the onion (Allium cepa) and fed to male Wistar rats for 20 days (in concentrations equivalent to 0.5g or 1g per day of fresh onion) raised testosterone levels and enhanced sperm motility. Luteinizing hormone and sperm concentration increased at the higher dose, whereas sperm morphology and testis weight didn't alter at either dose. Diallyl sulfide appears (based on these rat experiments) to dose-responsively elevate levels of luteinizing hormone, but too high a level of luteinizing hormone can reduce testosterone levels via negative feedback.

More experiments on rats indicate that doses of crude garlic higher than 5% reduce testosterone in conjunction with raising luteinizing hormone. Yet more rat experiments found that a dose of 5% crude garlic induced apoptosis in testicular germ cells, which was more pronounced at doses of 10% and 15%.

Thus, a little bit of garlic (or any other Allium extract) is good, but too much is bad. Theoretically, it is difficult to determine the ideal dose from these studies on rats, but it is much simpler in practice, since a small amount sprinkled on food is tasty and pleasurable, whereas larger amounts are increasingly unbearable.

Fenugreek
Fenugreek contains steroid-like alkaloids and saponins (or "steroidal glycosides"). These include diosgenin, smilagenin, sarsasapogenin, furostanol-type steroids and gitogenin. Oral ingestion of fenugreek steroids by diabetic rats not only reduced blood glucose levels but also upregulated testicular steroidogenic enzymes and elevated plasma testosterone (and estradiol).

Mucuna pruriens (Velvet Bean)
Three months of supplementing with 5g per day of powder derived from the plant Mucuna pruriens (velvet bean), by infertile men striken by psychological stress, increased sperm counts and motilities and increased seminal plasma levels of ascorbic acid and glutathione. On another occasion, also with infertile men, it did a similar thing and also raised levels of testosterone, luteinizing hormone and the catecholamines (dopamine, noradrenaline and adrenaline). The changes in catecholamines are hardly surprising, since the velvet bean contains their immediate precursor L-dopa, which (unlike the catecholamines themselves) is able to cross the blood-brain barrier and enter the central nervous system.

Serenoa repens (Saw Palmetto or Permixon)
An assortment of plant extracts and herbal products are promoted as inducing hormonal effects that are (or are erroneously believed to be) favourable for various populations. Claimed effects typically involve increasing total and free testosterone and/or lowering DHT and/or estradiol.

Serenoa repens (also known as permixon or saw palmetto) contains fatty acids in free form, primarily lauric (saturated, 12 carbon chains) and oleic acid (monounsaturated, 18 carbon chains). Under "in vitro" conditions, lauric acid inhibits the type 1 and type 2 isoenzyme of 5-alpha reductase (responsible for converting testosterone into DHT), and oleic acid inhibts the type 1 isoenzyme. Because of this, Serenoa repens is predicted to reduce DHT levels, even though its fatty acids are likely to be esterified (bound inside phospholipids or triglycerides) in an "in vivo" environment, thereby preventing the effect. Lauric acid can be quickly converted to palmitic acid (the most abundant free fatty acid along with oleic acid) "in vivo", and palmitic acid (along with stearic acid, also an abundant free fatty acid) inhibits neither isoenzyme.

Use of Serenoa repens (saw palmetto) supplements has been linked with acute pancreatitis in a case study of a 65-year-old diabetic using it to "treat" benign prostatic hyperplasia, and with pancreatitis and hepatitis in a case study of a 55-year-old reformed alcoholic using it with the same end in view.

Tribulus terrestris
Another plant extract, Tribulus terrestris, is heavily promoted for increasing testosterone levels, although 500mg three times per day for two days fails to increase the urinary ratio of testosterone to epitestosterone over the limit defined as legal by WADA, much to the disappointment of two female athletes who had attributed use of Tribulus terrestris to their positive drug tests.

On several occasions, Tribulus terrestris has been included along with Serenoa repens as part of experimental supplement complexes also including DHEA (the precursor to the sex hormones) and androstenedione (a direct precursor to both testosterone and estrone) and chrysin, indole-3-carbinol and, in one instance, gamma-linoleic acid (GLA). These supplements were intended to raise total and free testosterone levels while reducing (or at least not affecting) DHT and estradiol levels. The actual effect was not quite as planned!

One such supplement did raise free testosterone levels when given to middle-aged men for four weeks, and it also raised DHT and estradiol (both contrary to predictions). On another occasion, again on middle-aged men, it had no effect on total testosterone and raised free testosterone only marginally, but elevated DHT by 71% and estradiol by 103%! A very similar product also containing GLA did much the same thing.

When given to young males who were also performing resistance training three times per week, an identical or near-identical product (named "ANDRO-6") had no effect on total and free testosterone levels but elevated androstenedione, estrone and estradiol (perhaps it should have been named "ESTRO-6")! Muscle strength was increased equally in both the supplement and the placebo group, indicating that the hormonal effects of the product had no additive benefit in the context of their training regimen.

It is very likely that the effect of the above supplements on sex hormone levels was attributable to the presence of DHEA and especially androstenedione rather than some useless, overhyped plant extracts. By itself, 10mg or 20mg per kg of bodyweight per day of Tribulus terrestris (700-1400mg per day for a 70kg person), given to healthy males aged 20-36 for four weeks, had no effect on levels of androstenedione, testosterone or luteinizing hormone. Elite male rugby players increased strength and fat-free mass while taking 450mg per day for five weeks, but the placebo group fared equally well, indicating that "structured resistance training" and not the plant extract was responsible for the effects.

Stinging Nettle
The root of the stinging nettle (Urtica dioica) contains various lignans (e.g. 3,4-divanillyltetrahydrofuran) which (along with their metabolites as produced in human intestines) are able to bind to SHBG and (under "in vitro" conditions) dose-responsively inhibit the binding of SHBG to its receptor on human prostatic membranes. Of course, SHBG is more able to bind to cellular receptors (and subsequently bind sex hormones) when unliganded, and (contrary to the impression gained from "in vitro" experiments concerning DHT and cancerous prostate tissue) it is by no means undesirable to have androgens transported to cells under optimal physiological conditions.

A combination of Urtica root and Serenoa repens (saw palmetto) extract reduced symptoms of benign prostatic hyperplasia (BPH, not actually a disease) as effectively as finasteride (a medication which inhibits DHT production via the type 2 isoenzyme), and the subjects were less able to tolerate finasteride. Whether this is worth caring about is very much open to debate, and the case studies cited above concerning Serenoa repens, pancreatitis and hepatitis also need to be considered.

Mint (Mentha) and Other Lamiaceae Plants (e.g. Basil, Thyme, Sage, Marjoram)
A "spermicide" concoction including Mentha citrata oil in addition to other ingredients prevented (in most cases) the migration of sperm into the cervical mucous of women who had been selected for having high mid-estrus cervical mucous scores. It also prevented pregnancy in female rabbits. The multi-component nature of this concoction makes it impossible to assess the independent impact of mint extracts, but other experiments have looked at Mentha properties in isolation.

A mere 10mg or 20mg per day of Mentha arvensis leaf "petroleum ether extract" reduced the ability of male albino mice to fertilize normal females, despite maintained libidos and no other evidence of toxicity. Sexual organs were reduced in mass and certain sperm parameters were abnormal.

An orally administered 10mg per day aqueous extract from the leaves of Mentha arvensis had no effect on the sexual behaviour of male albino mice of proven fertility, but it impaired their fertility, reduced the mass of their sexual organs and resulted in abnormal sperm and reduced sperm concentrations, motilities and viabilities. The alterations seen after 60 days returned to normal after 30 days of discontinuation.

An aqueous extract of Mentha spicata Labiatae, or spearmint, promoted oxidative stress in the hypothalamus (indicated by reduced levels of enzymes involved in endogenous anti-oxidant defence) and reduced androgen receptors and the expression of steroidogenic enzymes and steroidogenic acute regulatory protein (STAR), and reduced sperm densities. It was suggested that the dose of spearmint (not mentioned in the abstract) impaired hypothalamic production of luteinizing hormone (LH) and follicle-stimulating hormone (FSH). So, what is the likely effect of ingesting small quantities of mint along with roast lamb, garden peas, Guernsey butter, sea salt and LH-stimulating crude garlic? Goodness knows ...

Herbal teas containing extracts from Mentha spicata (spearmint, 20g or 40g per litre) and Mentha piperata (peppermint, 20g per litre) elevated luteinizing hormone and follicle-stimulating hormone levels in male albino rats, but reduced testosterone levels and did some peculiar things to the testicles.

In 21 hirsute women (many with polycystic ovarian syndrome, PCOS), drinking two spearmint-infused teas per day for five days lowered free testosterone levels and increased those of LH, FSH and estradiol. What this indicates for the rest of the human population is far from clear, since hirsute women and those with PCOS are characterized by unusually high androgen levels.

Other members of the mint family (Lamiaceae, to which the stinging nettle is very closely related) have been observed to exert an effect similar to that of Mentha plants. A benzene extract from the leaves of basil (Ocimum sanctum), given at a dose of 250mg per kg per day for 48 days, did peculiar things to the sperm of albino rats. The same dose of the same extract given to the same or similar rats resulted in a complete failure to fertilize female rats. By contrast, an aqueous extract of Ocimum suave had no effect (including in high doses) on the fertility or other health parameters of Swiss mice.

An 800mg per kg aqueous extract or a 400mg per kg ethanolic extract of Salvia fruticosa (a type of sage), given to rats for 30 days, had no effect on the ability of males to impregnate females, but reduced the number of viable fetuses.

On the other hand, a volatile oil extracted from Origanum majorana (marjoram) minimized the testicular toxicity and diminished steroidogenesis that resulted from ethanol-induced oxidative damage (10ml per kg per day for 10 weeks) in male rats!

Soy Products and Isoflavones
Soy products contain isoflavones which lower testosterone levels. Four weeks of replacing meat protein with soy protein (tofu) lowered the free androgen indexes of a group of men aged 35-62. Three months of daily supplementation with 60mg of soy isoflavones by males aged 30-59 resulted in unchanged total testosterone levels but elevated SHBG and reduced free testosterone and DHT.

Consuming 40g of isoflavone-rich soy protein isolate per day for six months resulted (in males aged 50-85 and deemed to be at high risk for prostate cancer) in elevated estradiol levels and (in biopsied prostate tissue) reduced amounts of androgen receptors. As discussed in the vitamin D3 section, vitamin D3 antagonizes cancer in cultured prostate cells through a mechanism that is actually dependent on (or at least greatly enhanced by) DHT, and it increases the amount of androgen receptors while doing so.

Among the male partners of couples examined at an infertility clinic, estimated consumption of soy foods over the previous three months was negatively correlated sperm concentration (but had no relationship to sperm motility, sperm morphology or ejaculate volume).

Liquorice, Liquorice Root and Glycyrrhizin
Liquorice and its glycyrrhizin content have been reported to lower testosterone levels. 100g per day of liquorice (with 150mg of glycyrrhetinic acid) for nine weeks had a minimal effect on sex hormones in a small trial with members of both sexes. On the other hand, another small trial, this one with type-2 diabetic males (also afflicted by chronic hepatitis) given liquorice root containing 240-525mg of glycyrrhizic acid once per week for more than a year, found lower total and free testosterone concentrations and increased evidence of atherosclerosis.

Methylxanthine Alkaloids, Theobromine, Theophylline, Caffeine, Coffee, Tea, Cocoa and Chocolate etc.
As mentioned in the section on vitamin E, the vasodilator nitric oxide (which works synergistically with testosterone to induce erections) can be produced in response to both gamma-tocopherol and the polyphenol "anti-oxidant" epicatechin, both of which are present in large amounts in dark chocolate. Dark chocolate (and especially cocoa butter) is also high in palmitic and stearic saturated fatty acids, higher dietary intakes of which (as mentioned in the section on fats and carbohydrates) are usually associated with higher testosterone levels. Palmitic and stearic acid are the two fatty acids which, in the form of free fatty acids under "in vitro" conditions, do not inhibit the isoenzymes of 5-alpha reductase (responsible for converting testosterone into DHT). Palmitic and stearic free fatty acids are also the ones which, again under "in vitro" conditions, do not promote oxidative stress within sperm cells. Some of the other fatty acids from the cacao plant occur in the form of peculiar lipids (e.g. anandamide) that have been characterized as "cannabinoid" due to their similarity to those found in the cannabis plant, although they do not produce false positive tests for cannabis, also known as marijuana (the potential effect of which on things such as steroidogenesis and spermatogenesis seems to be rather different, as summarised further down). Another thing prevalent in dark chocolate (and especially in cocoa powder) is the methylxanthine theobromine, a purine "alkaloid" that is a metabolite (along with theophylline) of the caffeine found in coffee (and in dark chocolate and cocoa in smaller amounts). Methylxanthines have been implicated in enhancing sexual function, and dark chocolate thus has massive theoretical potential to raise both testosterone and penile levels!

In contrast to coffee with its high caffeine content, which tends to have a very slight vasoconstricting, blood pressure raising effect, cocoa and dark chocolate with their high theobromine content tend to be vasodilative (including in the coronary arteries of heart transplant recipients) and to lower blood pressure (if it's high to begin with). Whereas coffee with its caffeine ramps up catecholamine production and has the potential to raise cortisol levels, a 40g daily dose of dark chocolate for two weeks has lowered urinary cortisol and catecholamine concentrations (evidence of a less stress-governed energy metabolism) in a small-scale study on humans.

The blood-flow effect of cocoa and chocolate has traditionally been attributed to theobromine, but three weeks of daily intake of flavanol-rich cocoa with added theobromine modestly raised ambulatory and lowered central systolic blood pressure, whereas the same without added theobromine had no effect. Judging by this evidence, it seems that higher doses of regular cocoa would slightly lower both ambulatory (due to the flavanols or perhaps just due to reduced intake or absorption of other foods) and central systolic blood pressure (due to the theobromine). Coffee also contains numerous other interesting properties that relax the effects mentioned above, which are mostly attributable to caffeine.

Although chronically elevated levels of stress hormones is normally a bad sign, their acute effect is to enable a person to respond to stressful situations and to make a person feel more energetic - a feeling that, if acted on, is unlikely to lead to chronic elevations of the said hormones. When taken prior to resistance exercise by high-level rugby players, 800mg of caffeine had no effect on performance and was linked with elevated concentrations of both testosterone and (more so) cortisol during and shortly after the workout. These transient rises were similar to but somewhat sharper than what is often seen after resistance exercise without caffeine.

Under "in vitro" conditions in rat Leydig cells, green tea extract and one of its isolated components (epigallocatechin-3-gallate, EGCG) both reduced testosterone production, but another of its isolated components (epicatechin, one of the flavanols found in cocoa and dark chocolate) did not reduce it. It is not clear what effect these entities would have in real life, but the effect of the methylxanthines on sexual functions is reasonably well-studied, even if not in live humans.

Caffeine, theophylline and related xanthine compounds have successfully stimulated sperm motility in semen samples taken from healthy young men, but they have failed to achieve the same thing in samples from men at a fertility evaluation clinic (those who actually need it most). On another occasion, however, theophylline did enhance the capacity of semen from infertile men (and from fertile men) to penetrate zona-free eggs, albeit ones belonging to hamsters rather than humans! In isolated testicular interstitial cells, theophylline increased basal cyclic adenosine monophosphate (cAMP) production and that stimulated by human chorionic gonadotropin (hCG), and it increased basal testosterone levels and (at a low dose) had no effect on testosterone production stimulated by hCG, but it reduced hCG-stimulated testosterone production at a high dose.

Although they certainly have their uses, methylxanthines are foreign substances that have the potential to harm the body in high and purified doses. Consuming caffeine or theobromine at 0.5% of diet for 14-75 weeks resulted, in almost all of the rats tested (aged 4-6 months), in testicular atrophy and impaired spermatogenesis despite (especially in the theobromine group) elevated testosterone concentrations. The effect of theophylline was similar but somewhat less potent. Here we have another example of foreign substances elevating testosterone levels while also messing up some of the most crucial things for which testosterone is needed.

Consuming purified methylxanthines has more potential for harm than consuming them within dietary items such as chocolate or coffee. Over the course of a week, 500mg per kg of bodyweight of purified theobromine per day (equivalent to me eating 4500g per day of very dark chocolate, compared to my personal best of a measly 600g) impaired the functioning of rats' Sertoli cells, but the same quantity given within cocoa extract failed to have any observable toxic effect. Bring on the chocolate! (Let me add, however, that personal experience leads me to believe that you'd become very ill if you ate as much as 600g of dark chocolate every day, mainly due to the effect of its indigestible and anti-nutritive properties, including fibre, phytic acid, oxalic acid, polyphenols/tannins etc.)

Although methylxanthines within dietary items (especially chocolate) are generally perfectly safe for adult humans, it is possible that they have the potential to be harmful to fetuses and infants (the same as ethanol, but not so severely when comparing amounts typically ingested, although doses of methylxanthines matching typical ethanol intakes could well cause instant death). This is because fetuses and infants lack the enzymes necessary for eliminating them. In tentative support of this possibility, an ecological correlation has been found between the incidence of testicular cancer and hypospadias in various countries and higher estimated consumption of cocoa in those same countries during the fetal and infant life of the cases.

Other Alkaloids, Nicotine and Tobacco Smoke, and Recreational Drugs
Methylxanthines are purine alkaloids or alkaloid-like substances which can be addictive and which can be toxic in sufficiently large (especially isolated) doses. In this respect they are similar to both legal and illegal drug-like substances, including cigarette smoke (the alkaloid nicotine plus other, more addictive substances) and both stimulant and opioid drugs. An important difference is that methylxanthines are less addictive and need to be ingested in much higher doses before having a toxic effect on the body (gonads included).

Like ethanol, methylxanthines and other exotic substances, nicotine and other components of cigarette/tobacco smoke have the potential to harm fetuses short on enzymes. Exposure to maternal tobacco smoke was not found to effect the hormone levels and semen characteristics of young adult male offspring, but current smokers among them had a higher percentage of abnormal sperm.

Tobacco smoke can be added to the list of substances that can do peculiar things to the sexual organs while potentially elevating testosterone levels. Male current smokers had higher luteinizing hormone and testosterone levels, and more voluminous right testicles, but also had impaired penile velocities.

As with ethanol, the potential testosterone-raising effect of tobacco is more pronounced in women than in men. A correlation has been found between maternal smoking and testosterone levels and the smoking habits and testosterone levels (higher) of adult daughters. Higher androgen levels in females, if stemming from lower estrogen production, are often an indication of reduced fertility. Among young adult females, testosterone levels were correlated with cigarette usage and later onset of puberty.

Salivary cotinine (a metabolite of nicotine) was correlated with salivary (free) androstenedione and testosterone at baseline among a set of male smokers. Quitting smoking was associated with reduced salivary androstenedione (but not testosterone) one year later. In some cases, higher levels of testosterone in the serum could be merely an effect of lower levels where needed most. Levels of cotinine and one of its metabolites in the semen of infertile smokers were correlated negatively with sperm motility.

Cigarette smokers were found to have lower serum prolactin levels, and male smokers had higher serum testosterone but lower semen testosterone levels. High-nicotine cigarette smoking, like intravenous cocaine, acutely increased luteinizing hormone without affecting testosterone. It also elevated prolactin, contrary to the prolonged effect reported above and also contrary to the acute (but not chronic) effect of cocaine.

In concentrations estimated to reflect levels in the testes of male human smokers, nicotine dose-responsively inhibited the penetration of zona-free hamster eggs.

The above research on tobacco smoke speaks for itself. What about marijuana?

Oxidative damage is partly responsible for the adverse effect of cigarette smoke on sexual functions (despite sometimes superficially favourable effects on hormone measures). Smoke and burnt food of any kind are capable of promoting oxidative damage and carcinogenesis. Smoking of marijuana (also known as cannabis, weed, hashish etc.) was more frequently reported by men with organic erectile dysfunction, in whom insulin resistance and diminished endothelium-dependent vasodilation (quite contrary to the effects of cocoa and dark chocolate with "cannabinoid" fatty acids) were linked. It is difficult to assess the worth of trends from small-scale observational reports, especially regarding not-very-legal entities that are often obtained in adulterated form. The property in marijuana that is thought to be psychoactive is delta-9-tetrahydrocannabinol, a non-alkaloid.

Studies have found both diminished and normal sex hormone levels in regular cannabis users and in response to acute experimental intakes. It is reported that one marijuana cigarette or an intravenous infusion of delta-9-tetrahydrocannabinol is sufficient to depress testosterone levels (an effect predicted to last for 24 hours). Experimental smoking of one or two marijuana cigarettes significantly lowered luteinizing hormone (and non-significantly lowered follicle-stimulating hormone and total and free testosterone) and significantly elevated cortisol in a very small study of four males. By contrast, no alterations in various sex- or stress-related hormones (or lymphocyte function) were found among six males who inhaled 18mg of delta-9-tetrahydrocannabinol from a marijuana cigarette.

Ephedrine, the alkaloid from Ephedra plants from which semi-synthetic drugs are derived, on one occasion normalized the sperm count of a single azoospermic male. Other relevant research concerns only semi-synthetic derivatives of this alkaloid.

Human chorionic gonadotropin increases cyclic adenosine monophosphate (cAMP) in addition to testosterone in cultured rat testes, but the semi-synthetic ephedrine-derived alkaloid amphetamine (which also increases cAMP) inhibits testosterone production (except when a cAMP-reducing "adenylyl cyclase inhibitor" is added to the mix). It is also thought that amphetamine (also known as "speed") reduces testosterone production via reducing calcium channel activity and the activity of various steroid metabolizing enzymes. Mammalian spermatozoa have both beta-adrenergic (which stimulate cAMP production by adenylyl cyclases) and alpha2-adrenergic receptors (which inhibit it), a pattern which has been suggested to facilitate a potential positive effect of amphetamine on fertility, and which may or may not be connected to the fact that a significant minority of amphetamine mono-users report enhanced rather than diminished sexual desire - although the average score of this minority on something called the "International Index of Erectile Function" was still lower than that of control subjects.

One can only speculate on whether the sexual conquests of Lemmy from Motorhead, who modestly claims to have slept with a mere 1000 women despite reports of double that figure, stem from the confounding effect of administering Jack Daniels, Coca-Cola and nicotine alongside amphetamine.

Methamphetamine, also a derivative of ephedrine, impaired the ability of male mice to impregnate females at a dose of 15mg per kg of bodyweight, and had no effect at lower doses.

In adult male rats, doses of methylenedioxymethamphetamine (MDMA, ecstasy), chosen to represent typical human raver intakes of an acute nature (3mg per kg, such as have been observed to increase serotonin and dopamine concentrations in the nucleus accumbens) and a chronic nature (0.5mg per kg or 1mg per kg doses to an average daily total of 4mg per kg), decreased testosterone production via a decrease of hypothalamic gonadotropin-releasing hormone. By contrast, 12 weeks of three-consecutive-days-per-week MDMA/ecstasy intake (up to 10mg per kg) designed to simulate human weekend clubbing usage failed to affect measures of sperm morphology or motility or mating viability in more male rats, but did result in some DNA damage to sperm. As for hormone concentrations during the course of an actual session of clubbing, the self-administered MDMA/ecstasy intakes of healthy volunteers resulted (compared to a control session without drug use) in a 75% increase in testosterone but an 800% increase in cortisol!

Another stimulant alkaloid is cocaine, which is extracted from leaves of a plant (Erythroxylum coca) that has been used in whole (and safer) form by South American populations for aeons. Like ethanol and methylxanthine alkaloids, cocaine can pass through the placenta, and this results in lower testosterone levels in the amniotic fluid of male babies born to cocaine users. Male cocaine users have exhibited hyperprolactinemia, which is thought to suppress immune function and increase susceptibility to infectious diseases. By contrast, the acute administration of 2mg of cocaine per kg of bodyweight to healthy males has lowered prolactin and elevated luteinizing hormone and FSH without affecting testosterone. Cocaine has had a somewhat ambivalent, dose-related effect (mildly adverse at high concentrations) on cultured human semen samples attempting to penetrate zona-free hamster eggs or bovine mucus. 30mg of cocaine per kg of bodyweight injected subcutaneously into male rats produced testicular vasoconstriction from 15 to 60 minutes after administration. 15mg per kg of bodyweight given twice weekly to more male rats did not affect the pregnancy rates of fertile females to whom they were mated but did reduce the diameter of the seminiferous tubules and the germinal epitheliums (also suggestive of testicular vasoconstriction). The same dose (representing the typical single dose of a heavy human user), given daily to yet more male rats for 90 days, reduced testicular levels of reduced glutathione and increased testicular levels of the lipid peroxide malonaldehyde. Long-term cocaine use by humans has been associated with reduced sperm concentrations, reduced sperm motility and other abnormalities.

The effect of opioid (analgesic, narcotic) alkaloids on sex hormone production and sexual functions is entirely negative.

Male users of methadone (a synthetic opioid drug that is used to treat addiction to others) had reduced testosterone concentrations and marked impairments in testicular function, including increased sperm counts as a consequence of reduced "secondary-sex-organ secretions". Users of heroin (a semi-synthetic derivative of morphine) fell somewhere between methadone users and control subjects on most measurements. On another occasion, heroin and combined heroin and methadone users all had (despite normal testosterone and gonadotropin concentrations) abnormal semen (including asthenospermia in every case), compared to just under half of the methadone users (who appear to have been using it as part of some sort of semi-effective treatment).

The effect of these drugs on steroidogenesis and sexual functions comes about quickly and severely as a result of the concentration of the opioid in the bloodstream. In heroin addicts (mostly males with a few females) being monitored under controlled circumstances, there was a distinct inverse relationship between plasma heroin concentrations and those of testosterone and DHT.

The adverse effect of opioids on sexual organs (contrary to the likening of their effect to orgasm) is attributed to acute diminishment of luteinizing hormone secretion. Luteinizing hormone and total testosterone are not always severely affected, but sexual functions are and free testosterone is typically depleted. Follicle-stimulating hormone can also be depleted (the addicts in this study had testosterone levels barely one-third of those in the control subjects). Yet more addicts had reduced levels of the most active thyroid hormone (triiodothyronine, T3) along with reduced cortisol and testosterone levels. Chronic heroin users with severely depleted testosterone and luteinizing hormone levels also had reduced vertebral bone mineral densities compared to control subjects, whereas former users who had abstained for at least four (and up to 24) months gave measurements similar to those of the controls.

Use of opioids under medical supervision for pain relief can be just as hazardous to the sexual organs. Males and females who received (into their spinal chords in treatment for pain) an average of nearly 5mg of morphine per day for an average of over two years had (in most cases) reduced levels of various sex hormones and gonadotropins in addition to reduced cortisol measures (including diminished cortisol and growth hormone responses to hypoglycaemia in around 15% of the subjects). Still more alarmingly, all but one of the males had decreased libido or outright impotence, and all but one of the pre-menopausal females failed to ovulate!

We now turn from that which reduces vertebral bone mineral density to that which enhances it.

Strength, Power and "Anaerobic" Training
There is usually a moderate elevation of testosterone after a reasonably challenging training session. When the training is of a nature (i.e. a fairly high volume of work with fairly heavy loads, or an excessive amount of volume with any load or none) that induces significant damage to the muscles and connective tissues, this rise is accompanied by a rise in cortisol. Normally these rises return to baseline very quickly, but there have been examples of both highly trained and relatively untrained subjects showing an increase in basal testosterone concentrations after periods of training. In the section on protein and amino acids, I mentioned that sprinters (training with sprint intervals) and track and field power athletes both had higher fasting testosterone levels at the end of a training phase (in which they'd been taking in a mere 1.26g of protein per kg of bodyweight). I also mentioned that collegiate American football players (who were using creatine) had sported higher resting free testosterone concentrations after a ten-week phase of resistance training.

Resting free testosterone also increased in men aged about 30 (previously active by way of jogging and recreational sports) who undertook 10 weeks of three-days-per-week weight training using typical exercises. An examination of overnight hormone concentrations in trained weightlifters following evening training sessions found that they secreted more testosterone in the last phase of sleep than on nights following no training.

Although extended sessions of vigorous training have the potential to do lots of damage to the muscles and connective tissues, vigorous intermittent training of a short duration can faintly lower cortisol levels. Performing 5 15-second Wingate anaerobic power tests separated by 30-second rest intervals, in hot or "thermoneutral" (great word) conditions, has no discernable effect on testosterone or its ratio to cortisol, but lowers cortisol itself in the immediate post-exercise phase. By contrast, a single minute of non-stop vertical jumps is sufficient to elevate levels of total and free testosterone, SHBG, both thyroid hormones and their stimulator, and adrenocorticotropin hormone (ACTH) and cortisol! The testosterone elevations were more pronounced among those who could jump higher.

Another factor to consider, besides hormone responses, is the quantity and location of androgen receptors, and these have become more numerous on the thigh musculature in response to short-term quadricep-centric training (along with higher free testosterone). On the other hand, six sets of 10 squats has reduced androgen receptor content along with acutely raising free testosterone. This is not necessarily a cause for alarm. Damage to muscles and connective tissues during resistance training is one way of forcing the body to adapt to greater challenges, and it could be that reductions in androgen receptor content challenge the body to adapt by gradually accumulating more (and/or by secreting more testosterone and/or by increasing the affinity of the receptors).

4 sets of 10 squats with 90-second rest intervals using 75% of the subjects' 1-rep maxes raised both testosterone and cortisol for a short while; 11 sets of 3 reps with 5-minute rest intervals using 90% of their 1-rep maxes raised neither. Squat-induced testosterone elevations (with higher or lower reps using relatively lighter or heavier loads) can last as little as 10 minutes. I suppose, however, that such paltry and transient elevations could still be of some physiological significance, since the post-training environment could result in a heightened uptake of testosterone and other steroid hormones into muscles and bones.

In theory, adrenal testosterone surges during and after vigorous, challenging endeavours could be body's means of sparing lean tissue from oxidation and giving itself time in which to be fed under natural conditions in which creatine, BCAAs and other potentially beneficial supplements are not growing on trees. It would be interesting to see a training-with/without-supplementation trial that tests whether those with higher basal and training-stimulated testosterone levels in the unsupplemented group are at any disadvantage compared to those in the supplemented group.

National-standard weightlifters training after sleep loss have shown decreased post-training cortisol responses and no dips in performance. Training twice per day rather than once tends to raise both testosterone and cortisol levels (and the ratio of the latter to the former) a little. Cortisol increases in response to Olympic-style weightlifting may be beneficial. International rather than national level weightlifters, and those performing better on the day, were found to produce more cortisol during proper competition than during simulated competition. Their testosterone responses did not differ, and the testosterone levels of these lifters were no higher than those of sedentary subjects previously measured at the same laboratory.

Doing forced reps (only completed with help from others) of various leg extension exercises, rather than maximum reps with the given load, does not additionally increase the transient rise in free testosterone, but does heighten the cortisol response.

Performing a combined 50 sets of leg presses, squats, bench presses and lat pulldowns diminishes overnight production of luteinizing hormone and free testosterone, and elevates cortisol.

Hypohydration during resistance exercise (6 sets of back squats using 80% of 1-rep max, using any extra sets necessary to equal the total number of session repetitions performed during baseline testing) by seven healthy young resistance-trained males (each serving as his own control) increased cortisol and catecholamine levels (consistent with prior research using hypohydrated endurance exercise) and at 5% hypohydration (the more extreme degree) non-significantly lowered testosterone levels (inconsistent with prior research using hypohydrated endurance exercise). Post-exercise elevated glucose and free fatty acid levels suggested insulin resistance (which wasn't tested) via hypohydration-aggravated muscle damage from intense resistance exercise. Squatting performance tended to be diminished during the middle sets, but not on the first or last set (which I can readily believe).

Over 11 weeks of twice-per-week resistance training (in addition to sport-specific training sessions) using 10-rep or 6-rep maximum loads (or 80% of those loads for the parallel squat) for about eight exercises, those members of the Spanish national Basque ball team who performed 6 sets of 5 reps (with 10-rep-max loads) or 6 sets of 3 reps (with 6-rep-max loads) showed elevated total testosterone levels and lowered cortisol levels, whereas those who performed 3 sets to failure using the same relative loads (including additional reps with reduced loads after failing to complete a rep or pausing for more than a second) had a borderline-significant rise in cortisol. During an additional five weeks, both groups performed 3 sets of 2-4 reps using 85-90% of their 1-rep maxes, and by week 16 the trainers-to-failure had a significantly greater increase in the number of repetitions with 75% of their bench press 1-rep maxes (as they had after weeks six and 11), whereas the trainers-far-from-failure (whose elevated resting total testosterone levels returned to normal during this phase) exhibited more power (speed of movement) during squats using 60% of their 1-rep maxes (unlike at the earlier time points). Here we have another example of a training phase (moderate volumes of far-from-failure resistance exercise with fairly heavy loads in addition to sport-specific training) elevating basal testosterone levels. Note that the training-to-failure group actually trained beyond failure, doing additional "drop sets" with reduced loads immediately upon reaching failure with the previous load. Whether simply doing as many reps as possible with a given load and then resting would elevate cortisol is another matter. I'd guess not, since the protocol above only resulted in a rise of borderline significance.

"Aerobic" Endurance Training
Two hours of running at a moderate set level (65% of maximum oxygen consumption, VO2 max) transiently raises testosterone (and growth hormone and the catecholamines) and lowers cortisol. The "stress response" to exercise of increasing vigour begins with growth hormone and the catecholamines and proceeds to adrenocorticotropin hormone (ACTH), followed by testosterone and/or cortisol and then maybe beta-endorphin. Beta-endorphin is an endogenous opioid that causes people to feel good. Given the effect of exogenous opioids on the sexual organs, could it be that addiction to exercise resulting in persistent performance of excessive volumes has a similarly catastrophic effect?

An hour of "aerobic" exercise at 65% of VO2-max increased testosterone and cortisol; an hour of intermittent training with two-minute "anaerobic" bursts above 100% of the intensity of VO2-max raised testosterone and also raised cortisol in a more pronounced fashion than the "aerobic" trial, but all of the alterations lasted only 1-2 hours (which is actually longer than those usually seen after single weight training sessions).

As mentioned in the section above, five "anaerobic" bursts lasting only 15 seconds each showed a tendency to lower cortisol. Acute testosterone responses to intermittent running correlate with lactate responses after a max test, and cortisol responses to steady-state 80%-VO2-max training correlate with VO2-max. Middle-distance runners have higher testosterone responses to the former; marathoners have lower cortisol responses to the latter.

Trained males running at c. 70% of VO2-max show a testosterone raise after half an hour but a return to normal by two hours. Trained women doing the same show no testosterone raise until about the two-hour mark.

Working at 60%, 70% and 80% of VO2-max in consecutive 15-minute phases on a cycle ergometer or a treadmill raised free testosterone levels (immediately after exercise) in trained or untrained women in their follicular or luteal phases.

Rowing 6000m for time on an ergometer results in raised testosterone and cortisol levels at the end, but the testosterone levels return to normal within the first half an hour, whereas the cortisol levels remain elevated at that time point.

No evidence of reduced sperm counts or depleted sex and reproductive hormones was found among male marathoners in a comparison to lean, age-matched control subjects, and no spermic abnormalities were found in another examination of male endurance athletes - except for lower sperm motility among professional cyclists during competition, which was attributed to mechanical testicular trauma!

Other examinations have painted a darker picture, however. Despite testosterone elevations in response to injected human chorionic gonadotropin and to two hours of running at 72% of VO2 max, six highly trained marathon runners had a lower amplitude and frequency of luteinizing hormone pulses, and lower LH responses to exogenous gonadotropin-releasing hormone, possibly due to too much negative feedback from transient testosterone rises during daily prolonged exercise.

Two hours of running at 80% of max heart rate actually raised testosterone in marathon runners (after shorter and/or less intense sessions on the previous three days), but they had lower testosterone and luteinizing hormone levels to begin with, compared to untrained controls (who had higher testosterone responses to the same session).

Males undergoing training for a marathon in 6-, 5- and 7-month phases terminated by road races of 15km, 25km and 42km, with their training distances never equalling that of the race being prepared for, showed a distance-dependent decrease in testosterone during the races but actually increased testosterone concentrations during the training periods (the volume of which is not given in the abstract).

Marathon running by women lowered sex hormone levels, but apparently for under two hours after finishing, and DHEA-S actually increased towards the end and remained elevated for some time.

Testosterone increased immediately post-event in marathon runners who averaged 2 hours and 33 minutes (very good but not elite), and in middle-distance runners training for an hour and in competitive walkers racing over 20km. Ultra-marathoners covering 107km in one go had lowered testosterone levels.

Middle-aged non-elite but well-trained males exhibited a rise in cortisol and a drop in free testosterone after running a marathon "under the conditions of the classic Athens marathon", which both returned to normal sometime within a week (presumably without running any more marathons).

Male participants in the Cardiff Marathon, whose hormone levels were monitored at 4-mile intervals, had increasing testosterone levels that began to drop after 21 miles, and cortisol levels which peaked half an hour after the finish and which were suggested to impair testosterone secretion.

A marathon run by males of unspecified (in the abstract) training status reduced testosterone over the first and second post-race day, despite elevated LH on the first three post-race days, along with (some of the time) elevated noradrenaline and adrenaline (which catecholamines were suggested to mediate stress-impaired testosterone biosynthesis, cortisol being elevated only immediately post-race).

Among male athletes aged 25-40, the post-marathon slashing of testosterone (by more than half) and spiking of cortisol (by more than half) had returned to baseline by the following morning (20 hours after the race). However, most of the muscle-wastage enzymes that were monitored rose even higher in the recovery period once hormone levels had returned to normal.

19km and 42km races by elite male kayakers elevated cortisol (more so at the longer distance than expected from data on marathons of the same distance, perhaps due to the repetitive use of the elbow joint) and lowered testosterone, but the latter changed much less between distances and both hormones returned to normal within 18 hours.

At 4000m altitude, running a marathon additively raised cortisol above altitude-induced levels and lowered free testosterone levels that had resisted altitude, and the cortisol rise returned to normal within 24 hours whereas free testosterone only partially recovered.

Endurance-trained runners showed alterations in cortisol and beta-endorphins only during the first 33km of a 110km run, but testosterone dropped throughout the race and was accompanied by dropped luteinizing hormone by the end.

A 15-day 400km road race resulted in testosterone reductions of only 31% and increased the cortisol-to-testosterone ratio by 83% in male marathoners on the day after the race.

It is clear from the above that the length of a regular marathon is close to (and in most cases a little beyond) the point at which the adrenals are able to continue secreting testosterone, and that the after-effect of such challenges (sometimes measured in hormone concentrations and sometimes measured in muscle damage enzymes) can linger for many days. So, what about the long-term effect of repeated sessions of this nature?

60 weeks of treadmill running for two hours at 80% of VO2 max on five days per week reduced free testosterone, LH and FSH (and their response to gonadotropin-releasing hormone), and sperm density, motility and morphology among habitually aerobically exercising males aged 20-40. The hormonal (but not spermic) patterns were the same in a group doing the same at 60% of VO2 max, and all the adverse changes recovered in both groups during 36 weeks of low-intensity exercise.

Runners averaging 108km per week had lower free testosterone levels compared to those averaging half as many and compared to sedentary control subjects. Their total motile sperm counts and densities (and their sperms' capacity for penetrating bovine cervical mucus) were significantly lower compared to those of the sedentary controls (but presumably not compared to the 54km-per-weekers), and they showed lesser sperm motility and a greater quantity of immature sperm cells compared to both other groups. All groups had an average age in their late 20s.

A week of hiking c. 50km per day by men and women who ate nothing (except "natural products", perhaps meaning wild foods) resulted in elevated cortisol and catecholamine levels in both sexes and lowered testosterone in the men (indicative of reduced testicular but not necessarily adrenal production).

Reading through the above makes it very clear that excess endurance exercise is an enemy of the sexual organs. Marathons seem to take at least half a week to recover from, and longer distances (or near-daily sessions of lengthy and moderately intense endurance exercise) seem to put one in a perpetual state of recovery. Covering the same distances in a non-competitive fashion (not trying to challenge one's best time) would probably have a less severe effect. Covering shorter distances less frequently without always working at a very high intensity has the potential (as indicated by the research summarized in the earlier part of this section) to moderately raise testosterone levels. Note that, although single sessions of running for an hour or two at 60-80% of VO2 max or max heart rate have elevated testosterone compared to at the start, running at 60% or 80% of VO2 max for two hours on five days per week lowered testosterone levels and (at the higher intensity) impaired sexual functions. Since testosterone protects lean tissues from oxidation during stress, it is probably a bad idea to persist with any mode of exercise beyond the point at which testosterone concentrations peak.

Stress Steroids, CNS Aromatase and Serotonin, Impulsive Aggression and Adaptive Insulin Resistance
Research concerning the relationship between hormone levels and "aggressive" behaviour is confusing and (especially on the surface) apparently contradictory.

An association has been found, among 89 prison inmates, between higher levels of SHBG and what were defined as "aggressiveness" and "anti-social personality disorder". It was suggested that this link existed due to the link between these things and use of various intoxicants which happen to increase SHBG levels.

Among 30 healthy males, total testosterone and free androgen index (FAI) correlated with sensation seeking, boredom susceptibility and disinhibition, but SHBG and total testosterone correlated with assault, indirect aggression and verbal aggression.

In response to experimental harassment, men who displayed more reactive hostility also had higher levels of both testosterone and cortisol.

Men and women displaying more aggression had lower SHBG levels and higher binding potentials (in various brain regions) of the serotonin 5-HT(1A) receptors. (As a matter of interest, SSRI drugs, designed to prevent the re-uptake of serotonin from the central nervous system and thus to maintain higher levels of serotonin in the CNS, can produce sexual dysfunction that persists after discontinuation in some subjects.)

Co-elevations of salivary testosterone and cortisol following administration of an SSRI are linked with serotonin depletion and aggression; co-depletions of salivary testosterone and cortisol following the same are linked with serotonin elevation but likewise with aggression.

It is the aromatization of testosterone in certain CNS regions (the posterior hypothalamus, for example) that is linked with aggression (and its intensity) in conjunction with dropping CNS serotonin. In "western" (and perhaps just about all modern) societies, high-testosterone males are often (but not always) delinquent, aggressive, substance-abusing, sexually promiscuous and not mindlessly obedient to authority (*gasp*), and are more likely to be imprisoned for violent or sexual crimes rather than other crimes. High-testosterone females in prisons exhibit "aggressive dominant behaviour" but low-testosterone ones are "sneaky" and "treacherous". To take a broader view, however, the aromatization of testosterone into estradiol can also promote paternal behaviour in mammals, which in many species involves a strong element of protection (defensive aggression). Low-testosterone, high-cortisol individuals are aggressive, but they resort to sneaky, underhand, indirect aggression rather than open confrontation.

Serotonin depletion in the CNS (or low cerebrospinal fluid 5-hydroxyindole acetic acid, 5-HIAA, a CSF metabolite of CNS serotonin) is linked with depression, impulsive aggression (high-testosterone equals aggression and low-serotonin equals impulse) and suicidality in individuals with high CSF levels of free testosterone (who also tend to have depleted cholesterol levels, probably reflecting elevated metabolization). Such a depletion can occur as a result of failed attempts to achieve dominance ("hyper-responsiveness to aversive stimuli"), or perhaps (to view it in a more complex light) due to the impossibility of achieving dominance (or enforcing one's will) through means (i.e. direct, open, non-deceitful) that such individuals perceive as being honourable. The hypothalamus and amygdala are particularly linked with both testosterone and serotonin activity. Exogenous testosterone (but not DHT) has increased the density of 5-HT2A receptors in higher centres of rat brains, and estradiol (but scarcely testosterone) has done the same in regions with minimal aromatase activity (a very strong indication that it is testosterone-derived estradiol and not testosterone itself that is responsible for the effect). The 5-HT2A serotonin receptor is linked, at least in rat skeletal muscle, with insulin-independent glucose transport, which suggests the possibility that the use of serotonin in skeletal muscle could encourage the depletion of CNS levels - hence the "adaptiveness" (under population-dense conditions) of the insulin resistance referred to in the next paragraph.

The notion of a so-called Soldier-to-Diplomat transition within more recent human societies has been incorporated into a more complicated and complete alternative hypothesis to the formerly popular notion of "thrifty genes" as an explanation of the propensity of modern societies towards visceral obesity. "Soldier" types from the animal kingdom are characterized by a triumvirate of "autistic" traits: high testosterone, low cholesterol and low CNS serotonin (plus other features including peripheral insulin sensitivity and secretion of epidermal growth factor to regenerate beta cells in anticipation of fights and wounds). "Diplomat" types, by contrast, are said to be characterized by features of diabetes and/or the "metabolic syndrome": low testosterone, high cortisol and cholesterol, elevated serotonin signalling, peripheral insulin resistance (via chronically elevated hypothalamic serotonin signalling), systemic inflammation and delayed wound healing. Dominant creatures among other animal species apparently possess precisely those biochemical features that characterize aggressive, frustrated males in human societies.

The Stronger-to-"Smarter" behavioural switch hypothesis is the name given to the newer alternative to the "thrifty genes" hypothesis. So-called "r" strategists are characterized as producing many offspring in whom little care is invested (like looting soldiers of old); so-called "K" strategists invest a lot of care in few offspring and come to the fore when it becomes expedient for large portions of a society to switch from a muscle-dependent to a brain-dependent lifestyle (although such as switch need not necessarily make people "smarter", since intelligence is required to discern how best to strengthen the body, and how best to tactically and strategically crush one's foes in battle and war). Relatively insulin-independent tissues including the brain (and placenta and erythrocytes) get more nutrients during peripheral insulin resistance or hypoinsulinemia, which also diverts more nutrients through the placenta of pregnant women to feed their offspring. More maternal weight gain normally equals bigger babies, but diabetic mothers don't even need to gain weight to give birth to bigger babies. Potential post-reproductive-age degenerative diseases in consequence of chronic insulin resistance are deemed to be of no concern for the forces of sexual selection (which operate only during reproductive age), and the fact that whole-body insulin resistance reduces ovulation does not matter during transient bouts of insulin resistance around pregnancy. Hyperinsulinemia (especially in the brain/CNS, which has insulin receptors despite being relative insulin independent) has been linked with certain identified cognitive enhancements. Among mammals, greater body sizes and slower reproductive rates correlate well with longevity and (in many cases) also with a kind of partial insulin resistance through some mechanism or other. Correlations are reported between childhood growth rate and insulin resistance, insulin resistance being "adaptive" among "K" strategists under both over-nutritive circumstances (facilitating sneaky social manipulations aimed at hoarding abundant resources) and under-nutritive circumstances (sparing nutrients for the brain rather than wasting them on muscles that are too frail to have a fighting chance).

Music, Films, TV, Games, Sports and Miscellaneous Entertainment
Musically gifted females tend to have higher testosterone levels than other females, whereas musically gifted males tend to have very low testosterone levels. Listening to music (of an unspecified type) has lowered salivary testosterone in men but raised it in women, although listening to music does also typically lower salivary and/or plasma cortisol levels in subjects of both sexes, including after general anaesthesia. Listening to "new age" music around surgery has lowered cortisol levels; listening to music of one's own choice (AC/DC?) has lowered them even more. Playing stressful video games with built-in music, or watching violent films, raises salivary cortisol. The effect of violent films on testosterone was not recorded, but aggressive film scenes had no effect on testosterone in one study cited further down.

Although I found no research on the matter, I suspect that "violent" scenes of a heroic/rescuing nature would elevate testosterone levels, as would music that causes one to envisage such scenes. I would also predict that watching soap operas causes cortisol to soar. Exposure to polycyclic aromatic hydrocarbons in a PC games room, but not use of the PC games room by itself, has lowered testosterone levels in males, more so in teenaged ones than in those in their 20s.

Watching erotic or sexual films (but not violent ones) raises salivary testosterone; watching dental surgery being performed lowers it. Watching Mr Bean (or another humorous film) raises salivary testosterone (in elderly patients with atopic dermatitis); watching the weather forecast fails to do so.

But will violent films that are also humorous raise testosterone without elevating cortisol?


Close-fought games of Shogi ("Japanese chess") elevate both salivary testosterone and cortisol in both the winner and the loser.

Vanquishing one's foes at ice hockey in front of a home crowd elevates salivary testosterone more than doing the same away from home. Watching one's favourite sports team win has sometimes been reported in the media to elevate testosterone; watching them lose probably elevates cortisol. Just as giving too much of a shit about the fortunes of a bunch of strangers kicking a ball around can elevate cortisol, so giving too much of a shit about the fortunes of a bunch of strangers contesting a presidential election (against a near-identical bunch of other strangers) can do the same. This effect is called "socio-political subordination".

Tip-Top Tips to Top-Up Your Testosterone Tank

?Cholesterol is the precursor (via a long and complicated pathway eventually reaching pregnenolone, progesterone and dehydroepiandrosterone/DHEA) to the sex hormones
?Testosterone (an androgen) is metabolized into the androgen dihydrotestosterone/DHT (via the enzyme 5-alpha reductase) and into the estrogen estradiol (via the enzyme aromatase), and estradiol is interconvertible with estrone
?DHT promotes male-typical physical features and estradiol female-typical ones, but both (and estrone) promote bone density and strength in both sexes
?During fetal life, subsequent mental and physical development as a male is promoted in the central nervous system (CNS) by a surge of testosterone that is aromatized into estradiol, which "de-feminizes" the CNS
?In post-fetal life, the activation of androgen receptors by testosterone and/or DHT is primarily responsible for "masculinizing" behaviour, but estradiol can also promote certain male-typical behaviours (when its abundance occurs alongside the activation of androgen receptors thanks to the "co-localizing" effect of testosterone) as well as promoting "feminizing" ones (when the "co-localizing" effect of testosterone is relatively slight)
?Minus fetal exposure to estradiol and the subsequent activation of androgen receptors by DHT and testosterone (the latter of which also activates estrogen receptors and "co-localizes" the impact of androgens and estrogens and their receptors), the CNS is regarded as inherently "female"
?Both androgens and estradiol are important for maximizing aspects of sexual desire in both sexes, since males lacking the aromatase enzyme show an increase in libido (and bone mineral density) in response to exogenous estradiol, and females reporting a lack of libido often prove to have very low levels of free testosterone
?Sex hormones are less able to enter the CNS when bound by sex hormone binding globulin (SHBG), but being bound by androgen-binding protein (ABP) does not appear to have any effect on the passage of testosterone across the blood-brain barrier
?SHBG is able to bind as a ligand to its own cellular receptors, as well as binding sex hormones as ligands, and it is better able to bind to its receptors when not already binding a sex hormone, but is perfectly able to bind a sex hormone after binding to a receptor (thus it is possible that "free SHBG" has an importance of its own)
?Cortisol (a stress-related cholesterol-derived steroid hormone that is particularly responsive to lowering blood sugar levels) occurs in much larger quantities than testosterone (which is also secreted from the adrenal glands in response to stress) and must serve some vital functions; it often increases along with testosterone in response to challenging situations, but excessively stressful situations result in a massive spike of cortisol and a depletion of testosterone
?Aldosterone is a stress-related cholesterol-derived steroid hormone that is particularly responsive to lowering blood salt/sodium levels or especially high blood potassium levels
?Low-fat diets typically lower total and free testosterone levels and SHBG in middle-aged males, but the same is not necessarily so in younger males or in athletes who are better able to metabolize the inevitable high carbohydrate intake on a low-fat diet
?High fibre intakes typically lower total testosterone and SHBG but have little effect on free testosterone
?High-carbohydrate diets result in more "de novo lipogenesis" (the synthesis of new palmitic acid and cholesterol via acetyl-CoA in the liver); this can result (in diabetics and other people with chronic insulin resistance and glucose intolerance) in lowered levels of total testosterone and SHBG (but not free testosterone) probably stemming from hyperglycemia and oxidative stress, or (in athletes and other people who are mostly glucose tolerant and insulin sensitive) in extra free testosterone stemming from acute rather than chronic lipogenesis
?Diets which provide too little carbohydrate to replenish the amount of glycogen used by the glycolytic muscle fibres and the amount of glucose consumed by the brain, and which force the body to increase gluconeogenesis to prevent hypoglycemia, probably increase cortisol production at the expense of testosterone
?Replacing saturated fatty acids with polyunsaturated fatty acids (PUFAs) typically lowers sex hormone production; this is because PUFAs are vulnerable to oxidative damage (an enemy of sexual functions), especially in the form of isolated oils and when exposed to heat and air, and because PUFAs as free fatty acids (judging by "in vitro" experiments on cultured cells) inhibit the production of DHT from testosterone and are toxic to sperm cells
?Docosahexaenoic acid, the very-long-chained omega-3 PUFA found in fish oil, is probably vital to healthy sperm but is also the fatty acid which (in free form in cultured cells) promotes the most oxidative damage
?The saturated fatty acids most commonly found as free fatty acids in live creatures, palmitic and stearic acid, both promote the conversion of testosterone into DHT in cultured cells and have minimal toxicity to cultured sperm cells even in concentrations massively exceeding the normal range
?Despite the oxidative damage caused by PUFAs in cultured cells and by isolated polyunsaturated oils in live humans, eating intact nuts and seeds (which have a significant PUFA content alongside protective nutrients) often reduces markers of oxidative damage in live humans
?Testosterone production increases when previously stored energy is being utilized (as during exercise) but decreases in response to excessive volumes of exercise and in response to caloric deficits that are great enough to stimulate the catabolism of skeletal muscle amino acids for gluconeogenesis
?Testosterone production increases in response to stress but decreases in response to excessive or combined stresses (e.g. high volumes of exercise with pronounced caloric restriction, or either of these with hypoxia or hypohydration)
?Extra dietary protein or supplemental amino acids can protect against the catabolism of skeletal muscle for gluconeogenesis and can preserve testosterone concentrations during severe or combined stresses, but protein intakes of (to hazard a rough guess) 30% of calories or higher reduce testosterone production as a result of forcing the body to produce energy inefficiently
?Elevations of testosterone (and cortisol) following bouts of resistance exercise can last as little as 10 minutes and are not a cause for pant-wetting, but there are a few recorded examples of training phases increasing testosterone concentrations in both relative beginners and high-level athletes
?Testosterone levels typically increase for a short while following one to two hours of fairly intense endurance exercise, and they have also risen during not-so-long-distance training periods leading up to participation in some competitive very long distance events, but marathons and even longer-distance events typically cause a temporary dip in testosterone production (and potentially less temporary evidence of muscle damage), and long-term near-daily covering of long distances has impaired sexual functions as well as testosterone production
?Depleted concentrations of the anti-oxidants ascorbic acid (vitamin C) and endogenously produced glutathione go hand in hand with impaired sexual function
?Vitamin E and the trace mineral selenium both serve as anti-oxidants and both are probably essential for fertility, although high supplemental doses of either have the potential for toxicity, and there are other potentially very useful forms of vitamin E besides the form (alpha-tocopherol) normally found in supplements
?Deficiency of zinc leads to hypogonadism; consuming extra quantities of it has increased testosterone production in hard-training athletes as well as restoring sexual functions to formerly hypogonadic subjects
?Consuming extra quantities of calcium and magnesium has increased testosterone production in hard-training athletes
?The mineral iodine is essential for production of the thyroid hormones, which metabolize cholesterol (precursor to the sex hormones), activate receptors which belong to a family including those for the steroid hormones and fat-soluble vitamins, and promote spermatogenesis independently (but also inhibit it via negative feedback when over-expressed)
?Vitamin A in the form of retinal is essential for sexual functions, and supplementation with vitamin A has repaired sexual functions, but high-dose supplementation (especially with retinoic acid) in the absence of synergistic nutrients (e.g. zinc, iodine and vitamins D3 and K2) can reduce DHT production, induce hypothyroidism and potentially have other adverse effects
?In cultured prostate and breast cancer cells, DHT and estradiol are capable of promoting growth in isolation, but vitamin D3 or its metabolites have a potent inhibititory effect that is enhanced by the presence of DHT or estradiol (they decrease expression of the enzyme that deactivates the vitamin), and vitamin D3 increases the expression of the hormone receptors
?Vitamin K2 activates proteins that are crucial for bone health and for preventing soft-tissue calcification, and it serves as a ligand for the steroid and xenobiotic receptor (SXR), with which it appears to interact to antagonize the proliferation of cancer in liver cells
?Herbs, spices and plant extracts to take a moderate interest in include garlic/onion, fenugreek and Mucuna pruriens
?Herbs, spices and plant extracts to be wary of include mint, soy and liquorice
?A mineral ion to be wary of is fluoride (as found in most toothpastes)
?Useless herbal supplements to avoid wasting money on include Serenoa repens and Tribulus terrestris


?Avoid being musically gifted - unless you're Misha Koklyaev: