The disturbing worldwide progressive decrease in semen quality over the last 40-years has multifactorial interconnected origins, including environmental, toxicological, behavioral, and lifestyle-related issues. Nevertheless, the continuous search for new clues or potential agents and close interactions among human and environmental factors has been considered an urgent task by government, institutions, and scientists (24). Our study demonstrates one previously unrevealed aspect of marijuana (Cannabis sativa) consumption with deleterious effects on a specific male cohort seeking fertility evaluation in a specialized andrology setting. It is widely accepted that marijuana cannabinoids compete with endo-cannabinoids to occupy cannabinoid receptors (CB1 and CB2) along the hypothalamus-pituitary-testicular axis (25). To this day, a large body of data emphasizes the role of cannabinoids in controlling reproduction by modifying gonadotropin release, fertility, and sexual behavior (26).

Our data bring forward a significant correlation between marijuana use and lower serum estradiol (E₂) levels. The correlation also confirms a point between marijuana use and higher T/E2 ratios in the absence of any observed effect on testosterone levels. To date, conflicting data exist. While two studies based on the use of cellular models have reported that Δ9-tetrahydrocannabinol (THC), cannabinol (CBN), and cannabidiol (CBD), three major marijuana compounds, had neither pro- nor anti-estrogenic effect (27, 28), it was reported elsewhere that these same compounds might reduce E₂ levels via their effect on the sex steroid metabolizing cytochrome P450 enzymes CYP3A and CYP1B (29, 30). However, these same studies reported that marijuana smoke condensate (MSC) showed both estrogenic and anti-estrogenic effects assumed to be mediated by other marijuana compounds (27). It was proposed that phenolic compounds could be responsible for the estrogenic effect (27), while polycyclic aromatic compounds (PAHs) might account for the anti-estrogenic effect most likely via the estrogen (ER) and the aryl hydrocarbon (AhR) receptor pathways (31–34). It was also proposed that MSC could also interfere with E₂ biosynthesis via aromatase activity inhibition (28), which could agree with our observed increasing T/E₂ ratios in marijuana users. Although THC, CBD, and CBN were not linked to any aromatase inhibiting activity (28), cannabidiorcol (CBD-C1), cannabitriol (CBT), and cannabiripsol (CBR), other biologically active marijuana cannabinoids have this effect (35). A pro-estrogenic action of marijuana consumption was also suggested at the clinical and epidemiological level since it was reported to be associated with gynecomastia (36).

Several habits and lifestyle factors influenced prolactin as its serum concentration was significantly modified by marijuana use, tobacco smoking, and sedentary in infertile men. Concerning the effects of marijuana use on serum prolactin concentration, it has been suggested that the brain cannabinoid receptor (CB-1R) is expressed close to hypothalamic dopaminergic receptors and that the administration of THC provokes an increase in dopamine (DA) release (37). Through negative feedback, DA influenced prolactin secretion by the anterior pituitary (37). The interfering effect of marijuana on DA receptors could explain our data. However, the literature is still controversial, as we reported an increase in serum prolactin levels in marijuana users, whereas some other studies observed no change (7, 11, 38) while others report a decrease (39). These divergent findings will have to be further investigated. Although prolactin's physiological importance for male reproduction remains uncertain, high levels of this hormone have been associated with male infertility, probably due to a direct deleterious effect on spermatogenesis and the inhibition of pulsatile gonadotropin release (40). Concerning the influence of a sedentary lifestyle on serum prolactin levels, it was reported that physical exercise was associated with greater prolactin secretion by the anterior pituitary (41). Although the correlation we found in our study between sedentary and serum prolactin levels were not extraordinarily strong, it appears to be positive and not negative as expected. Regardless, it remains within the physiological range, which may interrogate its actual relevance.

With tobacco smoking, we recorded a weak negative influence on prolactin serum levels. The literature exhibits conflicting data with some reports identical to ours (42, 43), but others reporting increased serum prolactin levels in tobacco users (44, 45). The adverse effect proposed is that nicotine could stimulate the central dopaminergic tonus, inhibiting prolactin release by the pituitary gland (42, 43). In an experimental study using rat pituitary cell-line, nicotine inhibited the prolactin gene expression (46).

Concerning SHBG, our data suggests regular marijuana consumption was positively associated with SHBG serum levels in infertile men. BMI is considered the primary determinant of SHBG circulating levels (47), mainly via hepatic visceral fat (48). Because of that, SHBG is classically seen as a biomarker for such conditions as metabolic syndrome, type 2 diabetes, and cardiovascular diseases (49). A Danish study concerning 1,215 young men reported similar data; however, once they controlled BMI and tobacco use, serum SHBG increase was no longer significant (8). It was recently suggested that SHBG testing could help evaluate infertile men, especially in the situation of hypogonadism (50), as SHBG binds androgens and estrogens and contribute to their bioavailability (51).

In the current study, we found that marijuana did not influence serum testosterone levels in infertile men. Once again, the literature reports conflicting data with results similar to ours (10–12, 52), in other cases, a decline (7) or even an increase (8) in serum testosterone level among cannabis users. Moreover, it is worthwhile to mention that semen quality and hormonal levels could be poor biomarkers for male fertility, while time-to-pregnancy (TTP) might be more representative to couples' fertility status. In this context, a recent large retrospective study revealed that marijuana use was not associated with TTP for men and women (53).

Nevertheless, we must report some limitations of our study. As is the case for cross-sectional studies, the causal inference of many questions may remain unclear. Retrospective studies are designed to evaluate pre-existing data; therefore may be subject to a reasonable and acceptable bias. Some complimentary data such as the duration of involuntary infertility, sexual activities, female partner age are missing. Concerning the applied heterogeneous definition of “marijuana use,” it is hard to distinguish more robust user (3–5 times a day) from those light consumers because temporal and quantitative relationships were often difficult to assess, and marijuana users tend to minimize or hide their habit, considered by many as relaxing and un-harmful. Furthermore, the lack of association between semen parameters' abnormalities and any of the habits studied may be justified by the lack of a matched control group of fertile “healthy” men.

Marijuana use is associated with hormonal imbalance in a cohort of infertile men, significantly lowering E₂ serum levels while increasing prolactin serum concentrations. Further prospective studies with larger cohorts will be pertinent to corroborate our findings and improve the understanding of the underlying mechanisms that modify and regulate these perceivable endocrine changes. A warning notice should be addressed to marijuana consumers, particularly adolescents, young adults, and men in the reproductive age groups.