Most of the interest in evaluating the role of caffeine on reproductive performance in general and on fertility in particular can be traced back to a series of experiments in rodents which documented an increased frequency of fetal resorption and congenital malformations of the skeleton in rats receiving caffeine via oral intubation [9]. These experiments, however, have little relevance to human exposure given that the adverse effects of caffeine were only observed at doses equivalent to consuming 50–100 cups of coffee per day. In fact, experiments by the same and other research groups where caffeine was administered via drinking water and at exposure levels more relevant to humans failed to identify any reproductive hazards of caffeine [10, 11]. Physiologic mechanisms linking caffeine to female reproductive function are also unclear. Some evidence suggests that caffeine may inhibit ovulation or corpus luteum function [12–14] and caffeine has been associated with higher early follicular phase E2 levels [13, 15]. However, caffeine has not been associated with indicators of ovarian aging [16] and, on the other hand, it has been shown to improve insulin sensitivity [17, 18], which can result in improved ovulatory function in women with polycystic ovary syndrome [19] and possibly healthy women.

Nevertheless, the reputation of caffeine as a reproductive toxicant was reinforced by the findings of the first report on the relation between caffeine and fertility in humans. Wilcox and colleagues used data from a study aimed at identifying risk factors for early pregnancy loss to examine the relation between caffeine and time to pregnancy among pregnancy planners [8]. They found that consuming the equivalent of one cup of coffee per day resulted in a 50 % decrease in conception rate per cycle and that there was a dose response effect of caffeine consumption and fecundability. Although this study had major advantages, including its prospective design and its restriction to pregnancy planners, it was not originally aimed at assessing the effect of caffeine on fertility. Therefore, investigators did not assess caffeine intake in detail and failed to obtain data on potential confounding factors [8]. Furthermore, the study evaluated 104 women who did not conceive during their first 3 months of trying, but excluded 117 women who did conceive during that time period.

Since this report an additional 19 studies have evaluated the relation between caffeine and female fertility. While ten of these studies have found a relation between coffee or caffeine intake and decreased fertility [8, 12, 20–27], most of them are retrospective [12, 20–25, 27]. A problem with retrospective studies evaluating this relation is that, since caffeine is almost universally considered to be a risk factor for infertility in the general population and is endorsed as such in leading reproductive medicine textbooks [28], retrospectively assessing caffeine intake is all but guaranteed to result in differential recall of actual intake between fertile and subfertile women thus creating a spurious positive association between caffeine intake and fertility. In fact, detailed reviews of the literature regarding the reproductive effects of caffeine have shown that retrospective studies and studies of lower methodological quality are more likely to identify caffeine as a risk factor for adverse reproductive outcomes, including infertility [29, 30]. When only the 8 prospective studies are considered, the majority [5] show no relation between caffeine and fertility [31–35] and one study even shows slightly higher fertility among caffeine consumers [36]. Only two prospective studies suggest a deleterious effect of caffeine on fertility [8, 37]. The largest and most recent of these prospective studies found no association between caffeine intake and risk of ovulatory infertility among 18,555 women followed over 8 years [34] and no relation with fecundability, the per-cycle probability of conception, among 3,628 women planning a pregnancy [33] (Fig. 4.2).

The relation between alcohol intake and fertility is equally cloudy. Alcohol has been found to induce a rise in estrogen resulting in decreased FSH secretion, impairing folliculogenesis and ovulation [13]. Animal studies indicate that alcohol may also have an acute effect at the level of the hypothalamus, inhibiting LH secretion and thus disturbing ovulation [38]. On the other hand, and like caffeine, alcohol intake has been linked to improved insulin sensitivity [39, 40] and is not linked to markers of ovarian aging [16]. As was the case in the caffeine literature, most of the 18 studies [27, 31, 34–37, 41–52] that have examined the association between alcohol intake and fertility in women have reported deleterious effects of alcohol. However, many of the methodological issues that plague the literature on caffeine and reproductive hazards also apply to the literature on alcohol, starting with study design issues; only seven of the studies conducted to date have been prospective [31, 34–37, 47, 48]. Among the prospective studies, three report decreased fertility with increasing alcohol intake [31, 37, 48], two report no association between alcohol and fertility [34, 36], one found decreased fertility with higher alcohol intake among women older than 30 years of age but a similarly strong association in the opposite direction among younger women [47], and one reported significantly decreased fertility among slow acetylators and no relation among rapid acetylators [35].

While a careful review of the existing literature on the relations of alcohol and caffeine with fertility does not provide compelling evidence that they significantly hamper fertility, the opposite belief is deeply ingrained in the general population and among healthcare providers. Despite much research, it is not possible to draw strong conclusions regarding the role of alcohol and caffeine on human fertility and further study is clearly needed. Because randomized trials of moderate caffeine or alcohol consumption in relation to fertility may be judged as unethical by many, large prospective observational studies of pregnancy planners, preferably in populations with different patterns of alcohol and caffeine use, are necessary to determine whether moderate consumption of these substances affects fertility.

Dairy foods are an additional dietary factor that has been considered as a potential reproductive toxicant. Lactose, the main carbohydrate in milk, is cleaved in the intestine into glucose and galactose. In animal experiments, rodents fed a high amounts of galactose have decreased ovulatory rates and develop premature ovarian failure (POF) [53, 54]. This observation led to the hypothesis that high intake of milk and dairy products may increase the risk of infertility due to ovulatory dysfunction in otherwise healthy women [55]. However, few studies have been conducted in humans [50, 55, 56] and their results are not consistent.

Cramer and colleagues found a positive correlation between per capita milk consumption and age-related decrease in fertility rates in 31 countries [55]. On the other hand, a case-control study following up on these findings [50] found that women who consumed three or more glasses of milk daily had a 70 % lower risk of infertility when compared to women who did not consume milk [50]. A subsequent prospective cohort study found no relation between total intake of dairy foods and risk of ovulatory infertility [56]. However, this study reported unexpected associations between intake of reduced fat dairy foods with higher risk of ovulatory infertility as well as an association between intake of full-fat dairy foods with lower risk of this condition [56]. A closer inspection of the animal models also suggests that diets with galactose contents that are closer to relevant intakes in humans do not result in POF or other signs of ovarian damage [57]. Collectively, these data suggest that galactose is unlikely to be an ovarian toxicant at the usual intake levels of humans. While the epidemiologic literature suggests that dairy foods may influence fertility, this literature is nascent and requires further replication.

Clues that micronutrients may influence human fertility have been hiding in plain sight for decades, as case reports and case series, in the literature [58–66]. Many, but not all [58, 59], of these reports are among women with celiac disease, a condition known to be associated with a higher frequency of infertility and micronutrient deficiencies, most commonly of iron, folic acid, vitamin B12, and vitamin D [66, 67]. A growing body of literature suggests these micronutrients may be important in female reproductive physiology and fertility.

The strongest evidence is for the role of folic acid and other nutrients involved in its metabolism. Folate-requiring reactions, collectively known as the one-carbon metabolism, encompass a series of related metabolic pathways where one-carbon moieties are transferred from donors to intermediate carriers and ultimately used in methylation reactions or as building blocks in the synthesis of DNA [68, 69]. This pathway has a greater relevance when folate requirements are heightened due to increased demand for DNA synthesis such as gametogenesis and early embryo development [70, 71]. As early as the 1960s, it was shown in the immature superovulated rat that either an excess or deficiency of folates partially inhibited ovulation [7]. In rhesus monkeys, folate restriction results in irregular menstrual cycles, progressive depletion of ovarian granulosa cells, and decreased preovulatory serum estradiol and mid-luteal progesterone [72]. In a prospective cohort study of healthy reproductive-age women who did not consume diet supplements, women in the highest tertile of folic acid intake (median intake 271 μg/day) had 16 % higher luteal progesterone levels and were nearly 70 % less likely to have anovulatory cycles than women in the lowest tertile of intake (median intake 101 μg/day) [73]. In a separate prospective cohort study involving more than 18,000 participants, women in the top two quintiles of folic acid intake (median intakes 726 μg/day and 1,138 μg/day, respectively) were 40 % less likely to develop infertility due to anovulation than women in the lowest quintile of intake (median intake 243 μg/day) [74].

The effect of folic acid on ovulatory function may result from heightened ovarian responsiveness to FSH in high folate environments. In women undergoing controlled hyperstimulation with FSH, carriers of the T allele in position 677 of MTHFR (which leads to decreased enzyme activity and lower 5-methyltetrahydrofolate concentrations) have a decreased ovarian responsiveness to FSH, lower oocyte yield [75], and granulosa cells that produce less estradiol [76]. Similar effects have been found among carriers of the MTHFR A1298C variant [77]. In addition to its effect on ovulation, folic acid may further enhance fertility by improving embryo survival. In vitro studies of mouse preimplantation embryos show that endogenous folates are essential for embryo development due to its role in the synthesis of thymidine [71, 78]. Since thymidine does not accumulate in cells, accumulation of significant amounts of folates in the oocyte is required during gametogenesis to support the exponential increase in DNA synthesis that occurs during early embryo development [71, 78].

Iron status may also be important for ovulation and fertility as highlighted by studies regarding iron-transporting proteins in key ovarian cells. Transferrin (Tf) and its receptor (TfR) have been identified in granulosa cells and oocytes in several studies [79–81]. It has also been reported that granulosa cells can synthesize Tf which may be translocated to the oocytes [81]. Although it is possible that Tf and TfR are redundant in the ovary or do not play an important role in local iron metabolism, it has been suggested that these proteins are essential for ovum development and are required to support the increased iron demand of the developing follicle [80]. Additional evidence for a role of iron comes from studies documenting a higher risk of infertility among women with subclinical celiac disease. Undiagnosed celiac disease is more common among women with unexplained infertility than among fertile controls [66, 82]. Moreover, some of these infertile women have signs of iron deficiency including iron-deficiency anemia [66] and low ferritin levels without evidence of other nutrient deficiencies [82]. Intake of nonheme iron was more recently found to be related to lower risk of infertility due to anovulation in a large prospective cohort study [83]. Women in the highest quartile of nonheme iron intake (median intake 76 mg/day) had 40 % lower risk of infertility due to anovulation than women in the lowest quintile of intake (median intake 9.7 mg/day). Heme iron intake was unrelated to fertility [83].

The potential role of vitamin D on fertility has also received some recent attention. Physiologic and experimental data in animal models strongly suggest that vitamin D may play an important role in reproduction. Expression data shows that the vitamin D receptor (VDR) is present in the ovary [84, 85], the endometrium [84], and the placenta [86]. Vitamin D has also been found to stimulate the production of estradiol and progesterone in ovarian [85] and placental tissue [87], and to regulate the expression and secretion of hCG in human syncytiotrophoblasts [88] in vitro. Female rodents fed a vitamin D-deficient diet have reduced fertility [89, 90]. Knockouts for VDR and 1α–hydroxylase, which catalyzes the hydroxylation of 25(OH)D into the biologically active 1,25(OH)₂D, shed additional light into the role of vitamin D in reproduction. Female knockout mice have decreased fertility as a result of uterine hypoplasia, impaired follicular development, and anovulation [91–93]. Calcium supplementation partially reverses these reproductive effects in the knockout models [90, 94] and in the nutritional deficiency model [90], while vitamin D supplementation reverses them in the deficiency model [89], suggesting that there may be a combination of direct effects of vitamin D deficiency and effects mediated through secondary derangements in calcium and phosphorus homeostasis.

Human data on the role of vitamin D on fertility is less convincing. The three case-control studies that have evaluated vitamin D status in relation to PCOS provide mixed results. Li and colleagues found that vitamin D deficiency was nearly four times more common among 25 PCOS women than among 27 women with documented ovulation in Scotland [95]. This association was not adjusted for differences in BMI between these groups, however, and there were no statistically significant differences in PTH levels either [95]. Contrary to these results, Mahmoudi and colleagues found that 25(OH)D and PTH levels were significantly higher among 85 PCOS than among 115 ovulatory controls in Iran, independent of BMI [96]. A third study found no differences in the frequency of vitamin D deficiency between 93 PCOS and 71 controls in India [97]. Studies of vitamin D intake provide equally mixed results. Thys-Jacobs and colleagues supplemented 13 PCOS women with 50,000 IU of vitamin D weekly or biweekly plus 1,500 mg of Calcium daily for 6 months [98]. Regular menstrual cycles resumed in 7 of the 9 women with irregular menstrual cycles at baseline and two of them became pregnant [98]. Similarly, Wehr and colleagues supplemented 57 PCOS women with 20,000 IU of vitamin D weekly and found improvements in menstrual cyclicity in 30 % of women after 12 weeks of treatment and in 50 % of women after 24 weeks of treatment [99]. Nevertheless, a prospective study with more than 18,000 women found no evidence of a relation between vitamin D intake and risk of anovulatory infertility [56]. While vitamin D intake was strongly related to a lower risk of anovulatory infertility in age-adjusted analyses, this association disappeared upon adjustment for potential confounders, including BMI. A major limitation of this study, however, is that intake of vitamin D is a minor contributor to circulating levels of vitamin D [100]. In addition, vitamin D intake in the highest intake group (median intake = 783 IU/day) was substantially lower than in the single-arm trials suggesting a benefit of vitamin D supplementation. Therefore, lack of association with vitamin D intake in this study does not necessarily rule out an effect of vitamin D on anovulation-related infertility. Clearly, whether vitamin D plays a role on female fertility independent of its association with obesity remains an open question.

Dietary factors may also influence fertility through their effects on insulin and glucose metabolism. Insulin resistance, hyperinsulinemia, and hyperglycemia may be critical in the development of PCOS as demonstrated by the effect of insulin sensitizers and other antidiabetic medications on ovulatory function and fertility in these women [101–104]. These mechanisms may also influence ovulation and fertility among women with no clinical evidence of PCOS and those who do not meet all the diagnostic criteria [105–107]. For example, insulin resistance and postprandial hyperinsulinemia are more prevalent among oligomenorrheic non-PCOS infertile women than among eumenorrheic infertile women [105]. Likewise, diabetic women in the general population have lower fecundability than nondiabetic women [106]. Moreover, HbA1c levels within the nondiabetic range are inversely related to fecundability and positively related to characteristics of PCOS (free testosterone levels and cycle irregularity) among healthy pregnancy planners in the general population [107]. Since the quality and quantity of carbohydrates [108–111], the amount and sources of protein [112–115], and certain types of fat [116], are known to influence insulin sensitivity, it is possible that the macronutrient composition of diet could influence ovulatory function, and ultimately fertility.

The strongest evidence for this hypothesis to date comes from the Nurses’ Health Study II, which used data from more than 18,000 women and nearly 27,000 pregnancies and failed pregnancy attempts. Analysis from this cohort of women revealed that a 2 % increase in energy intake from trans-fatty acids (approximately 4 g of trans fats) at the expense of carbohydrates was associated with a 73 % greater risk of ovulatory infertility after adjustment for potential confounders, and risk more than doubled when trans fats were consumed instead of monounsaturated fats [117]. The amount and source of protein were also related to ovulatory infertility in this study. Women in the highest fifth of total protein intake had 41 % higher risk of ovulatory infertility [118]. However, while protein from animal sources was associated to higher risk of infertility, protein from vegetable sources was associated with lower risk of infertility. Increasing protein intake by 5 % of calories (approximately 25 g of protein) was associated with a 19 % higher risk of ovulatory infertility when protein came from animal sources but a 43 % lower risk when protein came from vegetable sources [118]. Dietary glycemic load, a summary measure of the amount and quality of carbohydrates in the diet [119], was also positively related to a higher risk of ovulatory infertility. Women in the highest quintile of dietary glycemic load had 92 % higher risk of ovulatory infertility than women in the lowest quintile after adjustment for confounders [120].

This hypothesis is also consistent with findings from other studies among women with PCOS. Intake of high glycemic index foods was higher among 30 PCOS women than among 27 control women [121]. In addition, a small crossover feeding trial among 11 PCOS women [111] consuming a low carbohydrate diet (43 % vs. 56 % of energy) led to changes that would be expected to result in improved reproductive and metabolic outcomes including a reduction in free testosterone levels of borderline statistical significance [111]. The Nurses’ Health Study II data appears to be at odds, however, with the findings of two small randomized trials comparing the reproductive effects of low protein (15 % of energy) vs. high protein (30 % of energy) weight loss diets among overweight PCOS women [122, 123]. The protein content of diet had no effect on menstrual function or androgen levels in these studies, although there were some improvements in menstrual cyclicity [122] and reductions in circulating androgens [123] as a result of improved insulin sensitivity due to weight loss. However, since weight loss has such a dramatic effect on ovulatory function in PCOS women, it is not surprising that protein intake does not have a measurable effect on reproductive parameters within the context of significant weight loss.

Characterizing the global effect of multiple dietary factors as summarized by dietary patterns can provide a more realistic estimate of the effect that changing diet could have on fertility by taking into account the possible multiple interactions between individual foods and nutrients. The relation between dietary patterns and fertility has been examined in two prospective cohort studies to date (Fig. 4.3). In the Nurses’ Health Study II, investigators generated a “fertility diet” score that ranked women according to their intake of eight dietary factors they had previously found to be independently associated to risk of ovulatory infertility [124]. The highest scores were assigned to women with high intakes of protein from vegetable sources, full-fat dairy foods, iron, the ratio of monounsaturated to trans fats and more frequent use of multivitamins; low intakes of protein from animal sources, dietary glycemic load, and low-fat dairy foods. This “fertility diet” score was strongly related to a lower risk of ovulatory infertility and infertility due to nonovulatory causes. Women in the highest quintile of the score had 66 % lower risk of ovulatory infertility and 27 % lower risk of infertility due to other causes than women in the lowest quintile of intake independently of age, parity, BMI, and other potential confounders [124]. Furthermore, the authors estimated that, assuming these relations were causal, 46 % of all infertility cases due to anovulation could be prevented by changes in diet composition alone [124].

Similar findings were reported among participants of the Seguimiento Universidad de Navarra (SUN) cohort, which follows university graduates in Spain. Higher adherence to a “Mediterranean pattern” diet, characterized by higher intakes of vegetables, fruit, fish, poultry, low-fat dairy, and olive oil, was associated with a lower risk of seeking medical help for difficulty getting pregnant [125]. Specifically, women in the highest quartile of intake of this dietary pattern had 44 % lower risk of difficulty getting pregnant than women in the lowest quartile. The findings of the two studies are generally consistent with each other, although full-fat and low-fat dairy foods were grouped into different risk profiles in the two studies. While these two studies strongly suggest that overall diet pattern have an impact on female fertility, further replication of these findings is warranted.

Just as in females, caffeine and alcohol have been some of the most extensively studied aspects of diet as potential determinants of semen quality. Although some studies suggest that caffeine intake may be related to higher sperm motility [131, 132], there is extensive evidence that caffeine intake is not related to semen quality parameters. A 2011 meta-analysis that combined data from 1,256 men across multiple studies found no relation between caffeine intake with sperm concentration, motility or morphology [133]. Moreover, studies not included in the meta-analysis also find little evidence of a relation between caffeine and semen quality [14, 133–136], including a single-center study of 2,554 young men in Denmark [136], larger than all other studies combined. Given the limited predictive value of semen parameters on fertility [127, 137–139], data on the three studies relating male caffeine intake to time to pregnancy among pregnancy planners is particularly important. Jensen and colleagues found that the male partner’s caffeine intake was associated with a lower probability of conception when they were also nonsmokers [26], but two similar studies found association between caffeine and fecundability [26, 36, 43].

While chronic alcoholism can severely affect the reproductive hormone axis and spermatogenesis [140], there is not strong evidence to suggest that moderate alcohol consumption affects semen quality or fertility. The previously mentioned meta-analysis [133], which included data from 6,465 men for the evaluation of the relation between alcohol and semen parameters, found that alcohol consumption was related to lower ejaculate volume but was not related to sperm concentration, total count, motility, or morphology [133]. Also, the majority of the studies assessing alcohol consumption in the male partner in relation to a couple’s time to pregnancy have not found evidence of a deleterious effect of alcohol [36, 43, 44, 141]. Moreover, when a negative effect of male alcohol consumption has been documented, it has been among heavy drinkers. A retrospective time to pregnancy study among 2,112 pregnant women in the United Kingdom found that heavy alcohol consumption by male partners (defined as >20 unit/week) was associated with a significant reduction in fecundity (twofold longer time to pregnancy), but moderate consumption was not associated with decreased fecundability compared to those who did not consume alcohol [27].

There is also an extensive literature on the role of antioxidants on male fertility. Unlike the literature on alcohol and caffeine, however, most of the literature is based on randomized trials [142–176]. A description of the trials conducted to date can be found in Table 4.1. A meta-analysis published in 2011, evaluated the effects of antioxidant supplementation in subfertile men on semen quality parameters and, importantly, on clinical pregnancy and live birth rates [177]. Based on data from three trials [145, 146, 148] which collectively included 214 couples, the meta-analysis found evidence of a statistically significant increase in live births comparing in couples where men were randomized to antioxidants vs. control [177]. Similarly, based on data from 15 trials [142, 146, 148, 151, 157–161, 164, 166, 167, 169, 171, 174] which collectively included 964 couples, the meta-analysis found evidence of a statistically significant higher pregnancy rate in couples were men were randomized to antioxidants [177]. The meta-analysis also found evidence of significant benefit of antioxidant supplementation on sperm motility after 6 months of treatment (10 trials, 963 men) but not after 3 months (10 trials, 514 men) or 9 months (3 trials, 332 men) of treatment, as well as an increase in sperm concentration after 6 months (6 trials, 825 men) and 9 months (3 trials, 332 men) of treatment but not after 3 months of treatment (7 trials, 320 men) [177].

While this meta-analysis of randomized trials presents evidence that appears solid at face value, there are still many gaps in elucidating the role of antioxidants in the management of the couple seeking fertility treatment as clearly signaled by the authors by rating the quality of the evidence from this meta-analysis as “very low” and qualifying the results as inconclusive [177]. Some of the problems with individual trials are summarized in Table 4.1. There are additional problems with the meta-analysis that are worth considering. First, the definition of an antioxidant was extremely broad which led to the inclusion of trials of nutrients that are not antioxidants [178, 179] or that could influence male reproductive function through other mechanisms completely unrelated to preventing oxidative damage such as by influencing sperm DNA production [155], as discussed below. Second, very few trials examined a single antioxidant and the antioxidant doses varied greatly across studies. In fact, there were no two trials that compared the exact same intervention among all the trials included in the meta-analysis [177]. In other words, although pooled together in a single analysis, these were all trials of different interventions. Therefore it is very difficult to attribute the beneficial effects observed to any one nutrient or combination of nutrients and it is nearly impossible to identify the minimal doses of individual antioxidant nutrients that could be reasonably expected to have a clinical impact. Another major concern, particularly for trials reporting live birth or clinical pregnancy as the outcome, is that dropout rates were relatively high and tended to be higher in the control group raising concerns about differential reporting of hard clinical outcomes among controls.

Some of the gaps from the clinical trial literature can be filled by some observational studies addressing the role of antioxidants on semen quality. While few, they have documented dose–response relations between intakes of vitamins C, E, and carotenoids with higher sperm concentration and motility [180–182] or the risk of oligoasthenoteratospermia [183], suggesting that these nutrients may explain some of the beneficial effects observed in the trials. However, there is clearly much research needed to identify which antioxidants or combinations of antioxidants have positive effects on male fertility and at what doses.

There is some concern that pro-estrogenic or anti-androgenic exposures, some of which may be delivered via diet, may affect spermatogenesis [184, 185]. Soy beans and soy-derived products are the main dietary source of isoflavones; plant-derived polyphenolic compounds with estrogenic activity. While generally considered to have a weak estrogenic activity [186–191], isoflavones can bind strongly to membrane estrogen receptors [192], exert nongenomic actions potentially deleterious to male fertility [193], and have been related to male reproductive disorders in mammals [194]. Human data on the relation between soy or isoflavones with male fertility is scarce and inconsistent. Mitchell and collaborators supplemented 14 young men with 40 mg/day of isoflavones for 2 months and found no appreciable changes in semen quality parameters or reproductive hormone levels compared to presupplementation levels [195]. Song and colleagues investigated the relation between isoflavone intake and semen quality in a group of 48 men with abnormal semen parameters and 10 men with normal semen parameters and found that isoflavone intake was positively related to sperm count and motility and inversely related to sperm DNA damage [196]. On the other hand, Chavarro and colleagues examined the relation between intake of soy and semen parameters among 99 male partners of couples seeking fertility treatment and found that higher intake of soy foods was associated with lower sperm concentration [197]. In agreement with these findings, a 2013 study of 609 idiopathic infertile men and 469 fertile controls in China by Xia and colleagues reported that urinary levels of isoflavones were related to lower sperm concentration, total count and motility, and higher odds of idiopathic male infertility [198]. This later study is important not only because it is larger than all the previous studies combined, but also because it addresses one of the major arguments offered against a potentially deleterious role of soy on fertility: that Asian diets include high amounts of phytoestrogens from soy foods without any apparent deleterious effect on fertility. The data on this field is still developing and future work is needed, particularly focusing on the effects on fertility rather than semen quality, as well as further work among populations with high intake of soy.

Meat and dairy have also been hypothesized to be vehicles for environmental estrogens under modern dairy farming [199] and livestock production practices [200]. Specifically, because commercial milk is mostly obtained from pregnant cows [199, 201] there is concern that pregnancy hormones in milk [202, 203] could have reproductive effects in milk drinkers. Also, because anabolic sex steroids are administered to meat cattle for growth promotion in the United States and other countries [204] and residues are present in meat products [203], there is concern of reproductive consequences [200, 205] of meat consumption in places where this practice takes place.

Data on the relation between dairy and meat intake on semen quality is growing and suggests that these concerns may not be unfounded. Dairy food intake has been related to decreased secretion of LH, FSH, and testosterone [206]. In addition, intake of full-fat dairy foods has been related to lower sperm motility and morphology among healthy young men in the United States and to greater risk of oligoasthenoteratospermia among fertility patients in Spain [207, 208]. Also, a case-control study in Iran found a positive relation of borderline statistical significance between total dairy food intake and risk of asthenospermia [209]. A fourth study among fertility patients in the Netherlands, however, found no relation between dairy food intake and semen quality [210]. Data is equally split on the relation between meat intake and semen parameters. One study found meat intake to be related to lower semen quality among fertility patients in Spain [208], while another study among fertility patients in the Netherlands did not [210]. High beef consumption during pregnancy has been associated with lower sperm concentration among male offspring 30 years later [205]. It should be noted that the European Union banned the use of sex steroids for growth promotion in meat producing cattle in 1989 [204, 211] and therefore the association observed in the Spanish study cannot be attributed to sex steroid residues in beef. Further work, particularly in the United States and other countries where this practice still exists, is needed.

As was the case for female fertility, there is also strong evidence that folic acid metabolism is important in spermatogenesis and male fertility. This metabolic pathway is involved in the synthesis of purines and thymidine which are ultimately used in DNA synthesis [68, 69], an essential step in spermatogenesis. One-carbon metabolism appears to be particularly active in the testes [212–214], and genetic [215] or pharmacologic [216–218] disruption of this metabolic pathway drastically affects spermatogenesis in animal models. In humans, genetic variation in this pathway has been related to semen quality. A 2007 meta-analysis on the association between MTHFR C677T and male factor infertility reported pooled odds ratios (OR) (95 % confidence interval (CI)) for male factor infertility of 1.39 (1.15–1.69) for TT homozygotes and 1.23 (1.08–1.41) for T allele carriers [219]. Also, a large study conducted in Korea reported an association between homozygocity for the variant G allele in MTR A2756G and nonobstructive azoospermia (OR (95 % CI) = 4.63 (1.40–15.31)) as well as an association between being a carrier (OR (95%CI) = 1.75 (1.07–2.86)) or homozygote (OR (95%CI) = 2.96 (1.51–5.82)) for the variant G allele in MTRR A66G and oligoasthenoteratospermia [220].

Intake of folic acid also seems to impact sperm production. In two trials of folate supplementation, blood folate levels had increased fivefold and seminal plasma folate had increased threefold, but no changes were observed in sperm concentration or motility in one study [221], while a 53 % increase in sperm concentration and a doubling in the proportion of motile sperm were observed in another study [222]. Similarly, in a randomized trial of folate, zinc, folate + zinc or placebo, subfertile men assigned to the folate + zinc arm had a 74 % increase in total normal sperm count compared to preintervention values and a 41 % increase when compared to post-intervention values in the placebo arm which did not reach statistical significance [155]. Folates from dietary sources also appear to have an impact on semen quality. In a study among fertility patients in Spain, men in the highest tertile of folate intake had an 87 % lower risk of oligoteratospermia than men in the lowest tertile of intake [183]. Likewise, seminal plasma levels of vitamin B12 and folate are positively related to sperm concentration [223, 224] and, among men who have previously fathered a pregnancy and have sperm counts above 20 × 10⁶/mL, seminal plasma folate is inversely related to sperm DNA fragmentation [225]. Furthermore, folate intake has been associated with a lower frequency of sperm aneuploidy [226]. Whether these effects on semen parameters have any impact on fertility is not known, however.