Marijuana Smoking and Markers of Testicular Function Among men From a Fertility Centre

Feiby L. Nassan; Mariel Arvizu; Lidia Mínguez-Alarcón; Paige L. Williams; Jill Attaman; John Petrozza; Russ Hauser; Jorge Chavarro; for the EARTH Study Team


Hum Reprod. 2019;34(4):715-723. 

In This Article


Men had a mean (standard deviation, SD) age of 36.3 (5.11) years and BMI of 27.5 (4.70) kg/m2. Most were Caucasian (88%), had a university degree (84%), and did not currently smoke tobacco (94%). Of the 662 men in our study, 455 (69%) provided one semen sample, 90 (14%) provided two samples, and 117 (18%) provided ≥3 samples. Most (88%) semen samples were analysed within 30 min after specimen collection and 72% of the men had a sexual abstinence of 2–4 days (Table I). Fifty five percent of the men reported having ever smoked marijuana; 44% of men were past and 11% were current marijuana smokers. Marijuana smokers were more likely to be white, overweight or obese and tobacco smokers. They also had higher intakes of alcohol and coffee and were more likely to have ever used cocaine (Table I). All but one of the men who reported ever use of cocaine also reported marijuana smoking. The distributions of the semen parameters, sperm DNA damage measures and hormone concentrations are shown in Supplementary Table II

Men who had ever smoked marijuana had significantly higher sperm concentration than men who had never smoked marijuana in unadjusted (Supplementary Table III) and multivariable-adjusted analyses (62.7 (95% confidence interval (CI): 56.0, 70.3) million/mL vs. 45.4 (38.6, 53.3) million/mL; P = 0.0003) (Table II). There were no statistically significant differences in sperm concentration between current and past marijuana smokers (P = 0.60). Similar patterns were observed for total sperm count. Men who had ever smoked marijuana also had 16% (−27%, −4%) lower serum FSH concentrations than men who had never smoked it, with no significant differences between past and current marijuana smokers (P = 0.53) (Table II). There were no associations of marijuana smoking status with other semen parameters, markers of sperm DNA integrity or other reproductive hormone concentrations. Of note, cocaine use was associated with a higher adjusted proportion of sperm concentration and count below the WHO reference values. In these analyses, marijuana smokers had an estimated 5% (95% CI: 3%, 9%) of semen samples with concentrations below 15 million/mL while never marijuana smokers had 12% (95% CI: 8%, 19%) (Figure I and Supplementary Table SIV).

Figure 1.

Adjusted means of the proportions (95% CI) of semen parameters below the WHO lower reference values associated with marijuana smoking status among 662 men (using the first semen samples given per man closest to assessment of marijuana smoking). Abbreviations: WHO; World Health Organization; 95% CI, 95% confidence interval. 1Adjusted mean proportions were estimated using generalised linear models with binary distribution and logit link. Models were adjusted for age (years, continuous), race (white/not), sexual abstinence time (days, categorical), body mass index (kg/m2, continuous), tobacco smoking (yes/no), coffee (binary) and alcohol intake (binary), cocaine use (yes/no), and calendar year (continuous). Motility models were further adjusted for time elapsed between semen collection and analysis. 2WHO lower reference limits (2010): ejaculate volume <1.5 mL; sperm concentration <15 million/mL; total sperm count <39 million; total motile sperm <40%; progressive motile sperm <32%, and normal sperm morphology <4% using 'strict' Tygerberg method. *P value < 0.05, **P value < 0.01 compared to never.

In analyses restricted to ever marijuana smokers, increasing marijuana smoking by 20 joint-years was associated with significantly higher serum concentrations of testosterone of 8% (2%, 15%), inhibin B of 11% (0.30%, 23%) and SHBG of 9% (2%, 17%) and a non-statistically significant higher sperm concentration of 12.6 (−6.55, 35.6) million/mL and total sperm count of 10.7 (−8.12, 33.4) million (Table III). In addition, a later age of initiation of marijuana smoking was associated with a non-statistically significant lower sperm count (−2.56 (−5.46, 0.42) million) and concentration (−2.94 (−5.84, 0.06) million/mL). Furthermore, each additional year since last smoking marijuana was associated with a 2.21 (0.13, 4.34) million higher sperm count, a 1.03% (0.16, 1.91) higher TDM and lower (−1.75% (−3.21, −0.27)) estradiol concentrations. The association between marijuana smoking and sperm concentration persisted after adjustment for serum testosterone (data not shown).

The associations between marijuana smoking status and markers of testicular function persisted after re-categorising exposure status based on last time of reported use, after restricting analyses to men without a diagnosis of male factor infertility, in analyses restricted to the first semen sample per man, after further adjustment for stress levels (Supplementary Table V, Supplementary Table VI, Supplementary Table VII and Supplementary Table VIII) or history of STDs, and after further adjusting for time of serum sample collection for testosterone (data not shown). A sensitivity analysis to quantify the impact of unmeasured confounding showed that in order for an unmeasured confounder to explain the observed relation between marijuana smoking status and sperm concentration, it would have to be associated with both sperm concentration and marijuana smoking status by a risk ratio ≥2.08 (or ≥1.59 to exclude the lower bound of the confidence interval) above and beyond the measured confounders.