Does Intermittent Fasting Lower Testosterone?
Small fasting trials found lower testosterone in trained men, but the evidence has limits. See what the studies measured and how to interpret a blood test.
A gym forum says 16:8 tanked someone’s testosterone. Another says fasting boosts it. Both cite studies. If you fast and lift, it’s easy to wonder whether your eating window is quietly working against you.
There are real hormone results behind the first claim. The leap from those results to a rule for every man is where the evidence gets thin.
The short version
Some small studies in trained young men found lower average testosterone during time-restricted eating. They do not show that fasting lowers testosterone in every man, or that a lower group average means symptoms or illness. The often-cited eight-week and year-long lifter papers overlap: the longer study followed a subset of the original participants. Your body size or fasting timer cannot predict your hormone result, and a diagnosis needs clinical assessment and consistently low measurements, not one number.
Total and free testosterone are different measurements
A total testosterone test includes hormone attached to proteins and the small amount circulating unattached. Free testosterone measures that unattached part. MedlinePlus’s guide to testosterone testing explains the distinction and why a clinician may request different tests.
That matters when comparing headlines. A study reporting a change in free testosterone has not necessarily measured the same outcome as a study of total testosterone. The units can differ too: the lifter papers below report total testosterone in nmol/L, while the cyclist paper reports free testosterone in pg/mL. Don’t compare the bare numbers across those units or tests.
Measurements also depend on when the sample was taken. Levels are usually highest in the morning. A result needs its laboratory’s reference range, the testing conditions and the reason the test was ordered; a screenshot of someone else’s report leaves most of that out.
What the eight-week lifter study actually tested
In Moro and colleagues’ 2016 trial, 34 men with at least five years of resistance training were randomly assigned to two meal schedules for eight weeks. Both groups followed the same supervised lifting program. Participants were asked to maintain their usual calorie intake; food diaries and dietitian interviews were used, rather than providing every meal.
The fasting group ate meals starting at 1 p.m., 4 p.m. and 8 p.m. Dinner could finish at 9 p.m., making the prescribed window 1 p.m. to 9 p.m. The comparison group started meals at 8 a.m., 1 p.m. and 8 p.m., with dinner also ending by 9 p.m. Calling 1 p.m. to 8 p.m. an eight-hour window confuses the last meal’s start with its end.
Average total testosterone fell from 21.26 to 16.86 nmol/L in the fasting group, versus 18.60 to 18.85 in the comparison group. The study reported a significant difference in change between groups. It also found greater fat loss with the shorter window, without a statistically significant difference in changes to fat-free mass or the tested lifts. That last finding is limited evidence, not proof that muscle or strength would be identical under either schedule.
The year-long paper was a follow-up
The 2021 paper from the same team followed volunteers from the earlier trial. Twenty-six agreed to continue; six withdrew, leaving 20 men for the year-long hormone analysis. One comparison-group participant also missed the body-composition and strength testing. These are overlapping participants, not two independent replications.
Its table gives total testosterone as 19.55 to 14.61 nmol/L in the eating-window group and 18.21 to 18.97 in the comparison group. The text reports a 16.8% reduction, but those table means imply about 25.3%. The paper does not reconcile that discrepancy. We show the table values below and do not use either percentage as a prediction.
By 12 months, the window group reported eating about 6% less energy than at baseline. Estimated arm and thigh muscle area decreased, while both groups increased their bench-press and leg-press performance. Muscle area was estimated from body measurements, not directly imaged. Neither the strength gains nor the hormone decline settles what happened to libido, fertility or long-term health.
One cohort · an eight-week trial and a year-long follow-up
Lower averages in the eating-window group
Average total testosterone at the start and the end, in nmol/L. Every row uses the same scale. These means do not show individual variation or confidence intervals.
Moro 20168 weeks · 34 men · usual intake prescribed
-
Eating window First meal at 1 p.m.; dinner by 9 p.m.
21.3 → 16.9 nmol/L (down 4.4)
-
Meals across the day First meal at 8 a.m.; dinner by 9 p.m.
18.6 → 18.9 nmol/L (up 0.3)
Moro 202112-month follow-up · 20 of the original men
-
Eating window First meal at 1 p.m.; dinner by 9 p.m.
19.6 → 14.6 nmol/L (down 4.9)
-
Meals across the day First meal at 8 a.m.; dinner by 9 p.m.
18.2 → 19.0 nmol/L (up 0.8)
Total testosterone, nmol/L (group averages)
The 20 men in the follow-up came from the original trial. These panels are overlapping evidence, not independent replications.
The dietary difference matters. This follow-up cannot cleanly tell us how much of the hormone change came from meal timing, eating less, losing weight or their combination. Its small, self-selected group of continuers also makes it a weak basis for predicting what every new faster will experience. Our article on fasting and muscle loss looks more closely at the body-composition question.
What other training studies add
In a 2020 trial of 16 elite cyclists, eight riders followed a restricted eating schedule and eight followed their usual meal distribution for four weeks. Free testosterone fell by about 27% from baseline in the restricted group. It also fell numerically in the comparison group. The paper reported a difference between groups at the final measurement, but not a statistically significant time-by-diet interaction for free testosterone.
Those are different comparisons. A significant fall within one group, together with a nonsignificant fall in another, does not by itself prove that their changes differ. The riders were also a very specific population in preseason training. The investigators did not collect food diaries during the intervention, so actual intake remained uncertain. The measured cycling performance outcomes did not show a clear deterioration; that does not establish hormone safety.
There is newer evidence too. The abstract of Gavanda and colleagues’ 2026 bulking trial reports no meaningful between-group hormone differences after 12 weeks of training with prescribed extra calories. We could verify the abstract, but could not access the full methods and results for this review. We therefore do not use it to calculate an effect or claim that fasting has no effect. Its existence is another reason to avoid saying all training studies show a fall.
Can body size tell you which way testosterone will move?
Not from these trials. Comparing lean lifters in one study with men with obesity in another also compares different training histories, diets, schedules and measurement methods. That is not a controlled test of whether body size changes fasting’s effect.
A 2026 Iranian trial randomized 20 previously sedentary men aged 20–30 with obesity. Both groups did supervised resistance training for eight weeks. One group also fasted before training, building from 12 to 18 hours; the paper does not describe a consistent daily 16:8 schedule in enough detail to reproduce one.
Testosterone increased in both groups. The reported between-group comparison was not statistically significant. With ten men per group and no detailed food-intake records, that result cannot prove equivalence, establish a testosterone boost from fasting or isolate weight loss as the cause. It is evidence against presenting the older lifter findings as a universal response, but it is not a rule that testosterone rises whenever a man with obesity fasts.
Read the sample before the headline
More papers do not always mean more independent evidence
34 trained men
Original eight-week trial
26 volunteered to continue.
6 then withdrew.
20 of those same men
Year-long hormone follow-up
Do not add 34 and 20 as if they were 54 different participants.
A separate study in men with obesity
10 + 10 men
Training + fastingTestosterone rose
Training aloneTestosterone rose
No statistically significant difference between groups. That does not prove equal effects or a benefit caused by fasting.
Why a pooled number does not settle this
A 2026 meta-analysis by Ghoreishy and colleagues reported no statistically clear pooled effect of fasting on total testosterone, with very low certainty, and a reduction in free testosterone based on two studies. Its total-testosterone estimate combines different populations, including women.
There are additional reasons for caution. Its included-study table lists both the original Moro lifters and their year-long follow-up, and labels the latter as involving both sexes even though the primary paper describes men. These issues make the review a poor basis for a precise, men-only estimate. We have not recalculated its pooled result. A nonsignificant average across unlike studies would not prove no effect or long-term safety even without those concerns.
The useful conclusion is narrower: some small trials detected hormone changes, but the evidence does not provide a reliable personal forecast. We also cannot infer a causal chain from a lower hormone result to a change in muscle, mood or sexual function when a study did not test that chain.
How to approach a blood test
The Endocrine Society’s clinical guidance recommends diagnosing testosterone deficiency only when relevant symptoms or signs accompany consistently low levels. It recommends confirming a low result with a repeat morning, fasting measurement. It also advises against routine screening of all men without a clinical reason.
Follow the preparation instructions from whoever orders the test. An overnight fast for a blood draw is a testing condition; it is not the same intervention as months on a restricted eating schedule. Extending your fast beyond the requested preparation is not a way to make the result more informative.
If you have a test booked, bring useful context: when you last ate, recent diet or weight changes, symptoms and the medicines or supplements you take. MedlinePlus advises discussing medicines with your provider; don’t stop them on your own. If one result is low, discuss the next step rather than comparing it with a lifter’s group average or choosing a supplement from a forum.
What to do with this
- Treat persistent symptoms as a reason to seek care. Low sex drive or erectile problems deserve a clinical conversation, whether or not you fast. A timer cannot tell you their cause.
- Check whether the window fits your meals and training. The year-long lifters ate less despite an instruction to maintain intake. If a shorter window makes your planned meals difficult, reconsider the schedule rather than assuming more fasting must be better. Planning protein across your meals is a separate nutrition task, not a treatment for a hormone result.
- Use logs for what they record. Atomic Fast can track fasting times and meals and show a daily protein target. Its phase labels depend on elapsed time, not a blood measurement. Neither reaching a phase nor meeting a logged target verifies hormone levels or nutritional adequacy.
- Keep the scale of the evidence in mind. A handful of trained men, a selected follow-up group and a small obesity trial cannot determine your individual response. Neither a reassuring average nor an alarming percentage replaces an assessment of your own situation.
Frequently asked questions
Does 16:8 lower testosterone? It lowered average total testosterone in a small trial of trained men and in a follow-up of some of those same men. That supports the possibility of a fall in those conditions, not a rule for everyone using 16:8.
Can intermittent fasting increase testosterone? The small obesity trial described here found increases in both training groups, without a significant difference between them. It does not establish fasting as the cause of the increase or as a testosterone treatment.
Does a lower result mean fasting is dangerous? A laboratory change alone cannot answer that. The studies have limited samples, follow-up and symptom assessment. Clinical guidance considers symptoms, repeated measurements and possible causes; these studies do not establish long-term safety or harm for every man.
Should I fast before a testosterone blood test? The Endocrine Society recommends morning, fasting measurements when evaluating suspected deficiency, with confirmation of a low result. Follow your clinician’s specific instructions and do not stop medicines without advice.
Does fasting affect testosterone in women? Women’s hormone studies address different clinical questions. Our article on fasting and the menstrual cycle discusses that evidence, including PCOS. The male training results here should not be transferred to those conditions.
Sources
Sources checked: November 17, 2026. Central trial methods and results were read in full. Access limits are stated below.
- Moro et al. (2016), Eight weeks of time-restricted feeding in resistance-trained males, Journal of Translational Medicine — Full text, methods and Tables 2–4 checked. 34 randomized men; eight weeks; meal completion by 9 p.m. Food intake was self-reported. Total-testosterone means and change comparisons support the first chart, not a personal prediction.
- Moro et al. (2021), Twelve months of time-restricted eating and resistance training, Medicine & Science in Sports & Exercise — Full text and Tables 1–3 checked. Follow-up subset of the 2016 cohort; 20 hormone participants, 19 for body-composition/strength tests. Table means conflict with the narrative percentage. Dietary intake and estimated muscle area limit interpretation.
- Moro et al. (2020), Time-restricted eating in elite cyclists, Journal of the International Society of Sports Nutrition — Full text and hormone results checked. 16 riders over four weeks. Within-group free-testosterone fall and endpoint comparison do not establish a significant difference in change; intake during treatment was not recorded in food diaries.
- Farahmand Khoshkebijari et al. (2026), Intermittent fasting with resistance training in men with obesity, Journal of Obesity — Full text, methods and hormone results checked. 20 men, eight weeks; testosterone rose in both arms, with no significant between-group comparison. Limited diet description and sample size; nonsignificance is not equivalence.
- Ghoreishy et al. (2026), Fasting, caloric restriction and reproductive hormones: systematic review and meta-analysis, Food Science & Nutrition — Full methods, included-study table and results checked. Mixed populations, very low certainty for total testosterone, overlapping Moro reports and an incorrect sex label limit use for men. No independent pooled reanalysis performed here.
- Gavanda et al. (2026), Time-restricted eating during a bulking phase, The Journal of Nutrition — Abstract only, checked against the authors’ university publication record. Full methods/results were unavailable for this review. Included as an access limitation and a counterpoint to claims of uniform findings; no numerical hormone effect inferred.
- Endocrine Society, Testosterone Therapy for Hypogonadism Guideline Resources (2018) — Official recommendations checked: symptoms/signs plus consistently low levels, repeat morning fasting measurement and no routine population screening. The linked journal guideline was not independently read in full.
- MedlinePlus, Testosterone Levels Test — Patient guidance checked for total/free tests, morning sampling, symptoms and medication preparation. Follow the ordering clinician’s instructions.
Atomic Fast product notes. Current iOS and Android source and product documentation checked for timer phase mapping, meal totals and saved nutrition targets. These are logging/display features, not hormone measurements or clinical assessments.
Atomic Fast records your fasting times and meals on iPhone and Android.
This article is for general education and isn’t medical advice. If you’re pregnant, managing a health condition, taking medication, have a history of disordered eating, or are considering fasts longer than 24 hours, talk to a healthcare professional first.