Vitamin D Raised Testosterone by 25 Percent in a One-Year Trial of 54 Men. The Same Research Group Ran Two More Trials in 198 Men and Found Nothing.
Ask ten men who have had their testosterone checked what they were told to do about it, and several will say they were told to take vitamin D. The advice traces to a single trial, published in 2011 by Stefan Pilz and colleagues in Graz, Austria, which gave 54 overweight men with low vitamin D levels either 3,332 IU of vitamin D3 a day or placebo for a year and found that total testosterone rose from 10.7 to 13.4 nmol/L in the supplemented group and did not move on placebo. A 25 percent increase from a vitamin that costs a few cents a day was an easy story, and it spread faster than the follow-up. Between 2013 and 2019, four more randomized trials tested the same question in 787 men, at doses equal to or higher than the original, in men who were healthy, men who were hypogonadal and men who were infertile. All four found nothing, and two of them came from the same Graz group that published the first result. This article walks through what each trial did, why the correlation that started the story is genuine even though the supplement does not act on it, what the 47,000-person outcome trials added, and what a man should actually do with a vitamin D result.
Key Takeaways: What the Vitamin D and Testosterone Evidence Actually Shows
The only randomized trial to report that vitamin D raises testosterone in men was published by Pilz and colleagues in Hormone and Metabolic Research in 2011: 54 overweight men with a mean 25-hydroxyvitamin D of about 30 nmol/L took 3,332 IU of vitamin D3 a day or placebo for one year, and total testosterone rose from 10.7 to 13.4 nmol/L in the 31 men on vitamin D with no change in the 23 men on placebo. Four later randomized trials in 787 men found no effect: Jorde and colleagues in Tromsø pooled 282 men from year-long trials at 20,000 to 40,000 IU a week, Heijboer and colleagues in Amsterdam pooled three trials, and the Graz group itself ran 98 healthy men in 2017 and 100 men with total testosterone below 10.4 nmol/L in 2019 at 20,000 IU a week for 12 weeks and found no change in total, free or bioavailable testosterone in either. A Danish trial of 307 infertile men given a 300,000 IU loading dose plus 1,400 IU a day for 150 days found no improvement in semen quality and no change in testosterone. In cross-sectional data, vitamin D and testosterone are consistently correlated: in 2,299 men in the LURIC cohort both peaked in August and troughed in March, and in 1,362 men in the Health Professionals Follow-up Study testosterone rose with 25-hydroxyvitamin D up to a plateau near 75 to 85 nmol/L. The two largest vitamin D outcome trials, VITAL in 25,871 Americans at 2,000 IU a day for 5.3 years and D-Health in 21,315 Australians at 60,000 IU a month for up to five years, found no reduction in cancer, cardiovascular events, fractures or all-cause mortality. About 5 percent of American adults have 25-hydroxyvitamin D below 30 nmol/L and a further 18 percent sit between 30 and 49 nmol/L, and correcting true deficiency remains reasonable for bone and general health. Vitamin D is not a testosterone intervention.
One Trial Moved Testosterone. Five Comparisons Did Not.
Reported difference in total testosterone between vitamin D3 and placebo, in nmol/L, across the randomized trials that measured it in men. Pilz and colleagues in 2011 gave 54 overweight men with a mean 25-hydroxyvitamin D near 30 nmol/L either 3,332 IU a day or placebo for a year and reported total testosterone rising from 10.7 to 13.4 nmol/L on treatment with no change on placebo, a difference of about 2.7 nmol/L. Every subsequent randomized comparison, including two from the same Graz group in 2017 and 2019 at 20,000 IU a week, the Tromsø pooled analysis of 282 men at 20,000 to 40,000 IU a week, three pooled Amsterdam trials and the Danish fertility trial of 307 men, reported no significant difference between arms, plotted here at zero. The null trials together enrolled 787 men against the 54 in the positive one.
Is vitamin D actually linked to testosterone in men?
Yes, in every large observational dataset that has looked, and the link is strong enough that the original hypothesis was reasonable. Wehr and colleagues reported in Clinical Endocrinology in 2010 on 2,299 men referred for coronary angiography in the Ludwigshafen Risk and Cardiovascular Health study. Men with 25-hydroxyvitamin D at or above 30 ng/mL, which is 75 nmol/L, had significantly higher total and free testosterone and lower sex hormone binding globulin than men who were insufficient or deficient, and the relationship held after adjustment for age, body mass index and other confounders. The same paper showed that both vitamin D and testosterone follow the same seasonal curve, with a peak in August and a trough in March, which is what a causal relationship would look like if sunlight drove vitamin D and vitamin D drove testosterone. Nimptsch and colleagues confirmed the association in 2012 in 1,362 men in the Health Professionals Follow-up Study, where total and free testosterone rose with 25-hydroxyvitamin D across the low range and then flattened at roughly 75 to 85 nmol/L. The biology looked supportive as well. Leydig cells in the testis express the vitamin D receptor and the enzyme that activates vitamin D, and male mice with the receptor deleted have low testosterone and impaired sperm production. A correlated hormone, a shared seasonal rhythm, a receptor in the right tissue and a knockout mouse with the predicted defect is the kind of case that gets a trial funded. It is also the kind of case that observational data produces for a great many things that fail when randomized.
The Shared Season That Looked Like a Cause
Schematic of the seasonal pattern reported by Wehr and colleagues in Clinical Endocrinology in 2010 from 2,299 men in the Ludwigshafen Risk and Cardiovascular Health study, with each hormone indexed to its own annual mean at 100. Both 25-hydroxyvitamin D and total testosterone peaked in August and reached their lowest values in March. The vitamin D swing is large, roughly 40 percent between trough and peak, because skin synthesis depends on ultraviolet B that is scarce at European latitudes in winter. The testosterone swing is far smaller, on the order of 10 percent. Two hormones moving in phase is consistent with one driving the other, and equally consistent with both responding to summer's extra daylight, activity, regular sleep and lower illness burden. The randomized trials settled which reading was correct. The curves shown are illustrative of the published pattern and not the exact monthly means from the paper.
- 25-hydroxyvitamin D
- Total testosterone
What did the trial that started the claim actually find?
The Pilz trial found a statistically significant increase in total, bioavailable and free testosterone after one year of vitamin D3 in a small group of overweight, deficient men, and no change on placebo. The men were a subset of a larger weight-reduction study run by Zittermann and colleagues, so everyone in the trial was following a structured programme aimed at losing weight while taking either 83 micrograms of vitamin D3 a day, which is 3,332 IU, or placebo. Fifty-four men had testosterone measured at baseline and one year. Mean 25-hydroxyvitamin D at entry was around 30 nmol/L, well inside the range most guidelines call deficient, and it rose to roughly 85 nmol/L on treatment. Total testosterone in the 31 vitamin D men went from 10.7 to 13.4 nmol/L, bioavailable testosterone from 5.21 to 6.25 nmol/L and free testosterone from 0.222 to 0.267 nmol/L, all significant. In the 23 placebo men, none of the three measures changed. The trial had real strengths: it was randomized, placebo-controlled and ran for a full year, and the participants were exactly the population in which a deficiency-correction effect ought to be visible. It also had the weaknesses that small trials always carry. The testosterone analysis was not the primary aim of the parent study, 54 men is too few to exclude a chance result, the baseline testosterone of 10.7 nmol/L sits at the low end of normal so regression toward the mean was in play, and the authors themselves described the finding as preliminary and called for a larger trial. Larger trials arrived. They just did not confirm it.
What the 2011 Trial Reported in Its 31 Treated Men
Total, bioavailable and free testosterone at baseline and after one year in the vitamin D arm of the Pilz trial, published in Hormone and Metabolic Research in 2011. The 31 men took 83 micrograms, or 3,332 IU, of vitamin D3 a day inside a structured weight-reduction programme, and their 25-hydroxyvitamin D rose from about 30 to about 85 nmol/L. Total testosterone went from 10.7 to 13.4 nmol/L, bioavailable testosterone from 5.21 to 6.25 nmol/L and free testosterone from 0.222 to 0.267 nmol/L, shown here multiplied by ten so it fits the same axis. The 23 placebo men, who followed the same weight programme, showed no significant change on any measure. The baseline total of 10.7 nmol/L sits at the bottom of the normal range, the sample was small, and testosterone was not the primary outcome of the parent study, three features the authors acknowledged when they called the result preliminary.
- Baseline
- After 1 year of vitamin D3
Why did the follow-up trials find nothing?
Four independent randomized comparisons between 2013 and 2019, in a combined 787 men, found no effect of vitamin D on testosterone, and the most likely explanation is that the original result was a chance finding in a small sample. Jorde and colleagues in Tromsø reported in Hormone and Metabolic Research in 2013 on 282 men drawn from placebo-controlled trials of 20,000 to 40,000 IU of vitamin D3 a week for six to twelve months. Serum 25-hydroxyvitamin D rose substantially and testosterone did not move. Heijboer and colleagues in Amsterdam pooled three of their own randomized trials in Clinical Endocrinology in 2015 and found the same. Then the Graz group ran its own replications. Lerchbaum and colleagues, in the Journal of Clinical Endocrinology and Metabolism in 2017, randomized 98 healthy men with 25-hydroxyvitamin D below 75 nmol/L and normal testosterone to 20,000 IU of vitamin D3 a week or placebo for 12 weeks and found no significant treatment effect on total testosterone, free testosterone, bioavailable testosterone or sex hormone binding globulin. In 2019, reporting in the European Journal of Nutrition, the same team randomized 100 men who actually had low testosterone, defined as below 10.4 nmol/L, to the same 20,000 IU a week for 12 weeks. That is the population most likely to benefit, and the result was again no significant difference from placebo on any androgen measure. When the group that published the positive result runs two more trials with tighter design, more men and a higher weekly dose and finds nothing in either, the field usually accepts the answer. The 2011 paper still gets cited far more often than the two that overturned it.
Fifty-Four Men Said Yes. Seven Hundred and Eighty-Seven Said No.
Number of men enrolled in each randomized trial that tested vitamin D3 against placebo and measured testosterone, coloured green for the trial that reported an increase and grey for the trials that reported no difference. The positive result came from the smallest trial. Lerchbaum and colleagues in Graz, the same institution that published the 2011 finding, ran 98 healthy men in 2017 and 100 men with total testosterone below 10.4 nmol/L in 2019, both at 20,000 IU a week for 12 weeks, and found no treatment effect on any androgen in either. Jorde and colleagues in Tromsø pooled 282 men from year-long trials, and Blomberg Jensen and colleagues in Copenhagen enrolled 307 infertile men for 150 days. The Amsterdam pooled analysis by Heijboer is not shown because its combined sample was not reported as a single figure. In every field of medicine, a positive finding from a small first trial that fails in larger replications from the same group is the standard signature of a chance result.
Does the dose or the starting level explain the difference?
Neither explanation holds up when the trials are laid side by side. On dose, the Pilz men took 3,332 IU a day, which is about 23,000 IU a week. Lerchbaum used 20,000 IU a week, Jorde used 20,000 to 40,000 IU a week, and all of the null trials raised 25-hydroxyvitamin D into the sufficient range, so a failure to deliver vitamin D cannot be the reason they failed to deliver testosterone. On starting level, the argument runs that the Pilz men were frankly deficient at about 30 nmol/L while the Graz replications enrolled men below 75 nmol/L, which includes many who were only mildly insufficient. That is a fair difference, but the Tromsø and Amsterdam cohorts included plenty of deficient men, the 2019 Graz trial selected for low testosterone specifically, and subgroup analyses restricted to men who began deficient have not produced a signal in any of the null trials. On duration, the original ran for a year while the Graz replications ran 12 weeks, and it is possible that testosterone responds slowly. Jorde's year-long data argue against that. The remaining candidate is the one the original authors flagged: 54 men, an analysis that was not the parent study's primary outcome, a baseline testosterone at the bottom of the normal range, and a placebo group of 23. Findings with those properties are the ones that replication most often erases, and this one was erased three times.
The Null Trials Did Not Underdose
Weekly vitamin D3 dose equivalent in international units for each trial's maintenance phase. The Pilz trial's 3,332 IU a day works out to about 23,300 IU a week. The two Graz replications used 20,000 IU a week, the Tromsø trials used 20,000 to 40,000 IU a week, plotted at the midpoint, and the Danish fertility trial used 1,400 IU a day after a single 300,000 IU loading dose. The two large outcome trials are shown in amber for scale: VITAL's 2,000 IU a day is 14,000 IU a week and D-Health's 60,000 IU a month is about 13,800 IU a week. Every null trial raised 25-hydroxyvitamin D into the sufficient range, so the absence of a testosterone effect cannot be blamed on a failure to deliver vitamin D. The Institute of Medicine's tolerable upper intake for adults is 4,000 IU a day, or 28,000 IU a week, marked by the dashed line.
Does vitamin D improve sperm quality or fertility?
The largest trial says no for semen quality, with a post hoc hint on live birth that nobody has confirmed. Blomberg Jensen and colleagues, whose earlier laboratory work had identified the vitamin D receptor in human sperm and testis, published a randomized trial in the Journal of Clinical Endocrinology and Metabolism in 2018 in 307 infertile men. The vitamin D arm received a single oral loading dose of 300,000 IU followed by 1,400 IU of vitamin D3 plus 500 mg of calcium daily for 150 days, and the comparison arm received placebo. The primary outcome was a change in semen quality, and there was none: sperm concentration, motility and morphology were no different between arms. Serum testosterone did not differ either. In a subgroup analysis the investigators had not pre-specified, men with oligozoospermia in the vitamin D arm had a higher live birth rate, 35.5 percent against 18.2 percent, but the overall live birth rate across all 307 men was not significantly different and a subgroup result of that kind is a hypothesis for the next trial rather than a finding from this one. The fertility literature has the same shape as the testosterone literature. Men with low vitamin D have worse semen parameters in cross-sectional studies, the receptor is in the right cells, and the supplement does not change the measurement.
What do the large vitamin D outcome trials say about longevity?
Two outcome trials totalling 47,186 people found that vitamin D supplementation did not reduce cancer, cardiovascular events, fractures or death, which is the same pattern as the testosterone trials at a hundred times the scale. VITAL, published by Manson and colleagues in the New England Journal of Medicine in 2019, randomized 25,871 American men over 50 and women over 55 to 2,000 IU of vitamin D3 a day or placebo and followed them for a median of 5.3 years. The hazard ratio for invasive cancer was 0.96 and for major cardiovascular events 0.97, neither significant. A 2022 report from the same trial in 25,871 participants found no reduction in total fractures, hazard ratio 0.98, in a population that was not selected for deficiency or osteoporosis. D-Health, published by Neale and colleagues in the Lancet Diabetes and Endocrinology in 2022, gave 21,315 Australians aged 60 to 84 a monthly 60,000 IU dose for up to five years and reported an all-cause mortality hazard ratio of 1.04, with no benefit on any cause of death. The meta-analyses that predate these trials found small mortality reductions concentrated in older institutionalized people given vitamin D3, and it remains plausible that correcting frank deficiency in frail populations does something the general-population trials could not see. What the general-population trials rule out is the idea that a man with a 25-hydroxyvitamin D of 50 or 60 nmol/L extends his life by pushing it to 100. That is the idea most vitamin D supplement sales rest on.
Why does the correlation exist if the supplement does nothing?
Because the things that raise vitamin D in a man also raise testosterone through their own routes, and the things that lower one lower the other. Vitamin D is made in skin exposed to ultraviolet B, so a man with a high level is on average a man who spends time outdoors, which means he is more likely to be physically active, less likely to be obese and more likely to sleep in line with daylight, and each of those independently supports testosterone. Body fat is the strongest confounder. Vitamin D is fat-soluble and sequesters in adipose tissue, so obese men have lower circulating levels for the same intake, and obesity also lowers testosterone through aromatase, insulin resistance and reduced sex hormone binding globulin. Chronic illness pushes both down together. The seasonal curve that looked like proof is explained the same way: summer brings more sunlight, more activity, more sleep regularity and less illness, and testosterone in men follows that rhythm whether or not vitamin D is on the pathway. Mendelian randomization, which uses genetic variants that raise lifelong vitamin D levels as a natural experiment, points in the same direction. Sutherland and colleagues reported in the Annals of Internal Medicine in 2022 on 307,601 UK Biobank participants and found that genetically higher vitamin D reduced mortality only in people who started below 25 nmol/L, with no benefit above that floor. A biomarker that predicts an outcome and a supplement that changes the outcome are different claims, and vitamin D is the clearest example in men's health of the first without the second.
How common is vitamin D deficiency, and who should test?
About 5 percent of American adults have 25-hydroxyvitamin D below 30 nmol/L, the level at which bone disease becomes a real concern, and a further 18 percent sit between 30 and 49 nmol/L, according to Herrick and colleagues' analysis of the 2011 to 2014 National Health and Nutrition Examination Survey. That leaves roughly three quarters of adults at or above 50 nmol/L, the threshold the Institute of Medicine set as adequate for bone health in 97.5 percent of the population. Deficiency clusters in predictable groups: people with darker skin, people who are obese, people who live at high latitude or work indoors, older adults whose skin makes less vitamin D, and people with malabsorption or on drugs that accelerate vitamin D breakdown. A man over 40 who is overweight, works indoors and lives north of the 37th parallel through winter has a fair chance of being below 50 nmol/L in March. The test is a serum 25-hydroxyvitamin D, it costs little, and it is the only way to know, since symptoms of deficiency are vague until it is severe. The reason to test is not testosterone. It is that frank deficiency contributes to bone loss, muscle weakness and falls in the elderly, and correcting it in a deficient person has evidence behind it that correcting it in a replete person does not. The Endocrine Society's 2024 guideline moved away from routine screening in healthy adults for exactly that reason, recommending testing where a result would change management rather than as a wellness measure.
How should a man over 40 use this evidence?
Check 25-hydroxyvitamin D once, correct a result below 50 nmol/L with 1,000 to 2,000 IU of vitamin D3 a day, stay under the 4,000 IU daily upper limit unless a physician is supervising higher doses, and take testosterone off the list of reasons to do it. A man whose level is already above 50 nmol/L gains no documented benefit from pushing it higher on any endpoint that has been tested in a randomized trial, and he should treat any supplement marketing that says otherwise as a claim without a control group. A man whose testosterone is low should look at the variables the trials do support: body fat, since losing 10 percent of body weight raises testosterone by roughly 2 to 3 nmol/L in obese men, sleep, since a week of five-hour nights lowers daytime testosterone by 10 to 15 percent in healthy young men, resistance training, alcohol, and sleep apnea if the snoring and morning headaches are there. If those are addressed and testosterone stays low on two morning samples, the conversation is about the hypothalamic-pituitary-testicular axis and whether enclomiphene, human chorionic gonadotropin or testosterone itself is appropriate, and that conversation belongs with a physician. Vitamin D still belongs in the picture for the same reasons it always did, which are bone, muscle in the deficient and the general prudence of not being deficient in anything. It never earned a place in the testosterone conversation, and the trials that were supposed to give it one took it away.
Frequently Asked Questions
Have more questions? Contact us or explore our full peptide intelligence platform.
Related Articles
Does Telomere Length Determine Lifespan? Short Telomeres Track With Earlier Death. The Genetics Say Longer Ones Raise Cancer Risk.
Does telomere length determine lifespan? Short telomeres track with earlier death, yet genetic studies show longer ones raise the risk of several cancers.
Low-Dose Aspirin Added 8 Percent to Male Mouse Lifespan. In 19,114 Healthy Older Adults, It Added No Healthy Years and Deaths Ran Higher.
Does low-dose aspirin help longevity? It extended male mouse lifespan 8%, but the ASPREE trial in 19,114 healthy older adults found no gain and more bleeding.
Is NAD+ a Peptide? No. It Is a Coenzyme Built From Two Nucleotides, and the Peptide Label Comes From the Clinic, Not the Chemistry
Is NAD+ a peptide? No. NAD+ is a coenzyme built from two nucleotides. What NAD+ is used for, what IV NAD+ studies found, and how it differs from SS-31.