Obesity and sleep disorders track with worse semen. Alcohol mostly does not.

The best-evidenced lifestyle associations with poorer semen quality are obesity, sleep disorders and smoking. Alcohol is not one of them: in the pooled data, drinking was associated with lower semen volume and worse morphology, but the effects on sperm concentration and motility both had confidence intervals crossing zero, and the whole set was graded very low certainty. That pattern — a handful of moderate-certainty findings surrounded by a much larger number of weak ones — is what the field actually looks like.

What the review did

An umbrella review published in the Asian Journal of Andrology searched Web of Science, MEDLINE and Embase for meta-analyses published from January 2000 to February 2025, using infertility, semen volume, sperm concentration, sperm count, morphology, motility and progressive motility as endpoints [s1]. Forty-three meta-analyses were included, covering 67 risk factors [s1].

Each of the 249 effect sizes it collected was graded with GRADE. The distribution is the review's most important single result: 136 (54.6%) were rated very low certainty, 59 (23.7%) low, and 54 (21.7%) moderate [s1]. Nothing reached high certainty. Every number below sits inside that constraint.

Weight

Obesity carried the most consistent signal across parameters, in a meta-analysis of tens of thousands of men. Standardised mean differences were −0.33 for sperm count (95% CI −0.44 to −0.21, 21 studies, 40,361 men), −0.33 for sperm concentration, −0.24 for progressive motility (95% CI −0.36 to −0.13), −0.22 for semen volume (95% CI −0.28 to −0.15), −0.22 for motility (95% CI −0.36 to −0.07) and −0.16 for morphology (95% CI −0.25 to −0.06) — all graded moderate certainty [s1].

Being overweight was a different matter. For sperm count the standardised mean difference was −0.06 (95% CI −0.11 to 0.00, 12 studies, 49,393 men); for concentration −0.04 (95% CI −0.09 to 0.02); for morphology −0.03 (95% CI −0.14 to 0.07); for motility −0.04 (95% CI −0.08 to 0.01) [s1]. Four of those intervals include zero. Only semen volume (−0.11, 95% CI −0.20 to −0.02) and progressive motility (−0.15, 95% CI −0.23 to −0.06) reached significance [s1].

The gap between the obesity and overweight columns is the useful finding. It is not a smooth gradient across body weight; the associations are concentrated at the obese end.

Sleep and smoking

Sleep disorders produced moderate-certainty associations across four of five parameters: a mean difference of −27.91 for sperm count (95% CI −37.82 to −18.01, 6 studies, 4,199 men), −5.16 for sperm concentration (95% CI −9.67 to −0.65), −2.94 for progressive motility (95% CI −5.28 to −0.59) and −0.52 for morphology (95% CI −0.80 to −0.24) [s1]. Semen volume showed nothing (−0.04, 95% CI −0.14 to 0.05) [s1].

Cigarette smoking gave low-certainty associations with sperm count (mean difference −8.92, 95% CI −12.40 to −5.44, 20 studies, 4,587 men), motility (−3.48, 95% CI −5.53 to −1.44) and morphology (−1.37, 95% CI −2.63 to −0.11), with semen volume not reaching significance (−0.16, 95% CI −0.33 to 0.01) [s1].

Alcohol

The alcohol meta-analysis, covering up to 16,395 men across 15 studies, is the clearest example of a widely assumed harm that the pooled data do not clearly show. Semen volume fell (mean difference −0.25, 95% CI −0.42 to −0.07) and morphology worsened (−1.87, 95% CI −2.88 to −0.86), but sperm concentration (−0.50, 95% CI −4.06 to 3.05) and motility (−1.53, 95% CI −3.84 to 0.77) both had intervals spanning zero [s1]. All four were graded very low certainty, with heterogeneity between 73% and 90% [s1].

That is not evidence that alcohol is harmless to fertility. It is evidence that the studies pooled so far cannot answer the question, which is a different and more honest statement than either "alcohol lowers sperm count" or "alcohol is fine".

What looked protective

Three exposures went the other way. Nut consumption was associated with better motility (standardised mean difference 0.51, 95% CI 0.24 to 0.78) and morphology (0.54, 95% CI 0.24 to 0.83), both moderate certainty, though from only two studies and 223 men [s1]. A healthy dietary pattern was associated with higher sperm count (mean difference 16.70 × 10⁶, 95% CI 2.37 × 10⁶ to 31.03 × 10⁶) and progressive motility (5.85%, 95% CI 2.59% to 9.12%), at low certainty [s1]. The review's summary lists regular physical exercise, nut consumption and adherence to a healthy dietary pattern as factors that may reverse declines in semen quality [s1].

Two internal problems in the paper

The review reports, in its narrative results, that being small for gestational age, of low birth weight and of very low birth weight were each associated with an increased risk of male infertility, giving odds ratios of 1.10 (95% CI 1.07 to 1.12), 1.16 (95% CI 1.05 to 1.27) and 1.75 (95% CI 1.23 to 2.49) [s1]. Its own supplementary table reports the same three exposures from the same source meta-analysis as 0.91 (95% CI 0.89 to 0.93), 0.86 (95% CI 0.78 to 0.94) and 0.57 (95% CI 0.40 to 0.81) — the reciprocals, pointing the opposite way [s1]. The paper does not state that it inverted them, and a reader cannot tell from the document which orientation the source intended.

Separately, the supplementary table reports the effect of a healthy dietary pattern on sperm concentration as a mean difference of 6.88 × 10⁶ per ml with a 95% confidence interval of 1.26 × 10⁶ to 1.26 × 10⁶ — identical bounds, which cannot be a confidence interval [s1]. We have taken the count and progressive-motility figures, which are internally consistent, and left the concentration figure out of the account above.

What a clinical guideline does with material like this

The AUA and American Society for Reproductive Medicine guideline on male infertility, amended in 2024, translates this evidence base into a small number of statements. It recommends concurrent assessment of both partners at the initial evaluation, and one or more semen analyses as part of the male evaluation [s2]. On the intervention most often marketed off findings like the ones above, it is direct: clinicians should counsel patients that the benefits of supplements such as antioxidants and vitamins are of questionable clinical utility in treating male infertility, and that existing data are inadequate to recommend specific agents [s2].

The distance between "obesity is associated with a standardised mean difference of −0.33 in sperm count" and "losing weight will let you conceive" is the distance this literature has not yet crossed. These are cross-sectional and observational associations, graded moderate at best, in a measure that varies enormously between men and within the same man over time.

This article is informational and is not medical advice.

Sources

• Current risk factors for male infertility and semen parameters: an umbrella review of systematic reviews and meta-analyses — Asian Journal of Andrology, 2026-01-13
• Diagnosis and Treatment of Infertility in Men: AUA/ASRM Guideline (2024 amendment) — American Urological Association, 2024-08-15