Phytoestrogen intake and other dietary risk factors for low motile sperm count and poor sperm morphology

Few potentially modifiable risk factors of male infertility have been identified, and while different diets and food groups have been associated with male infertility, evidence linking dietary factors including phytoestrogens and semen quality is limited and contradictory.


| INTRODUC TI ON
Male subfertility is a significant factor in approximately 50% of all couples experiencing a period of infertility. 1 Despite widespread concern about reported but disputed declines in semen quality and more specifically total sperm count, 2,3 few modifiable factors of semen quality have so far been identified (eg [4][5][6]. The characterization of these factors is of particular interest since they offer a means to improve male fertility and potentially reduce the need for assisted reproductive technologies. Diet is possibly one such factor, and while the relationship between conventional semen quality parameters (concentration, motility and morphology) and diet patterns and specific dietary components including dietary phytoestrogens has been studied (eg [7][8][9], the results from these association studies are often inconsistent. Increasing consumption of "Western diets" has been associated with increased sperm concentration, 10 but not in all studies, [11][12][13] and furthermore has been both positively 12 and negatively 10 associated with sperm morphology. Similarly, increasing intake of diets considered healthy (eg Mediterranean) has been reported to improve sperm concentration or total sperm count in some [12][13][14] but not all 10,15 studies. In contrast, sperm motility has not been associated with diets considered healthy 10,12,15 or indeed a Western diet, [10][11][12] though a "prudent" dietary pattern has been positively associated with motility 11 but not consistently so. 13,14 In addition, despite the reported associations between endocrine disrupting chemicals and male fertility, 16 there are few studies that have examined associations between semen quality and oestrogenic phytoestrogens that are found in plant-based foods (especially soy). Soy food intake has been inversely associated with sperm concentration but not sperm motility nor morphology, 17 but a feeding trial of a phytoestrogen supplement did not find any evidence of changes in semen quality despite increases in plasma levels of the isoflavone phytoestrogens, genistein and daidzein. 18 The inconsistency in the reported associations between these different food groups and semen quality may reflect methodological differences in study design (eg subfertile men vs healthy fertile men), study size or study populations (eg Asian vs Western). There are also intrinsic differences in the measurement of food intake and in how food groups were defined from one study to another. As a result, the evidence base linking diet to male subfertility is limited and often contradictory and there are no dietary recommendations currently reported in the UK National Institute for Health and Care Excellence guidelines on fertility problems: assessment and treatment. 19 To improve this evidence base, in this paper we examine associations between dietary factors (with a particular focus on phytoestrogen intake) and low motile sperm concentration and sperm morphology in a large multi-centre case-referent study (CHAPS-UK).

| Design and recruitment
The CHAPS-UK study was a multi-site, case-referent study, and the design and methods have been described previously. 20 Cases and referents were the male partners of couples attempting conception with unprotected intercourse for 12 months or more without success and were recruited between 1 January 1999 and 31 January 2002. Men were excluded if they had prior knowledge of their own semen quality, were not able to understand English, had had a medical condition (eg cystic fibrosis) or medical treatment (eg chemotherapy) that could have caused their infertility and either they or their partner had been sterilized previously. 20 Prior to their first clinic visit following recruitment, participants completed a short questionnaire on work, lifestyle and health factors and further information on these factors was obtained by interview at the clinic. Men were asked to complete a dietary questionnaire at home with their partner in order to obtain the most complete and accurate information.

| Semen analysis
Men were requested to abstain from ejaculation for a period of 3-5 days (depending on the clinic) prior to providing a semen sample for a diagnostic analysis. This sample was used not only for infertility investigations but also for this study, and it was analysed according to a protocol based upon the techniques outlined by the World Health Organization 21 as described previously. 5,20 Sperm concentration was estimated at an andrology laboratory associated with each centre using a haemocytometer. Motility was captured on videotape using a computer outstation commissioned for the study from Hobson Tracker Systems Limited, Sheffield, UK, and the tape was Conclusions: Dietary factors associated with semen quality were identified, suggesting that male fertility might be improved by dietary changes.

K E Y W O R D S
semen quality, sperm motility, sperm morphology, daidzein, phytoestrogens, diet Systems (Sheffield, UK). The machine was programmed to recognize as 'normal' stained sperm heads which fitted the dimensions given in WHO 1999 21 : a length of 4.0-5.0 mm and a width of 2.5-3.5 mm, with a length-to-width ratio between 1.50 and 1.75.

| Case definitions
Two case definitions were used, reflecting WHO guidelines. 22 The first was based on sperm motility, and cases were men with a motile sperm concentration (MSC) of <4.8 × 10 6 /mL: referents a MSC of ≥4.8 × 10 6 /mL. The second definition was based on morphology, and cases with poor morphology (PM) were men whose sperm showed <4% normal morphology; referents had ≥4% normal forms.

| Dietary exposures
Information on dietary habits was collected using a questionnaire (Appendix S1) that included a 65-item food frequency questionnaire (FFQ) covering the 12 months prior to recruitment to assess phytoestrogen intake that was developed from then available information on phytoestrogen levels in food (eg 23-25). The following exposures were assessed: • Type of diet (Q3.1; Appendix S1) coded into three categories: meat ± fish eater, fish eater, vegan or vegetarian.
• Quantity of butter/margarine spread on bread (Q5.4 Appendix S1) categorized as thick spread, medium spread, scrape, and none at all.
In addition, soy intake was estimated from 10 soy-based food questions (FFQ: Appendix S1). Soy intake (number of portions per day) was categorized into none, below median intake (<0.0506) and above median intake (>0.0506). The amount of daidzein and genistein consumed was estimated as described in Appendix S2 for those answering at least 60 FFQ questions (Appendix S1).

| Assessment of confounders
Confounders found to be significantly related to outcome in earlier analysis of these data 4,5,20 were re-examined here for those completing the dietary questionnaire (see Appendices S3 and S4) and all relating to outcome retained. Age was not related to motility or morphology. The factors retained for motility were ethnic group, testes surgery, manual work, wearing boxer shorts and abstinence (Appendix S3), and for morphology were BMI, cannabis use, season and abstinence (Appendix S4).

| Statistical methods
Analyses of the two outcomes (low motile sperm count and poor morphology) used unconditional logistic regression, accounting for clustering of subjects within the 14 fertility clinics, first without, and then with, adjustment for confounders associated with that outcome. Where more than one dietary factor was associated with the outcome (P < .10), the contribution of each factor was assessed using a Wald statistic. All analyses were carried out using the generalized linear latent and mixed models (gllamm) command within STATA 14.2 (Stata Corporation; Statcorp).

| Study population
A total of 2249 men were recruited and provided a semen sample and 1907 of these (84.8%) returned the dietary questionnaire. Sperm motility was determined for all 1907 men, and 440 were cases (23.1%) with a low MSC (<4.8 × 10 6 /mL). Sperm morphology was determined in 1673 men who had returned a dietary questionnaire, and of these, 274 men were cases (16.4%) who had poor sperm morphology (<4% normal forms). There was no relation between case status and likelihood of returning the questionnaire but those that did so were more likely to be older and of white ethnicity (Appendix S5).

| Risk factors for low MSC
In the fully adjusted model (adjusting for clustering and ethnicity, surgery, work, boxer shorts and abstinence: Appendix S3), low MSC was not positively associated with any dietary factor other than a

| Risk factors for PM
After adjustment for clustering, body mass index, cannabis use, abstinence and season (see Appendix S4), poor sperm morphology was positively associated with use of semi-skimmed milk and inversely associated with consumption of red meat more than three times per week (Table 3). When all other milk types and no milk (with similar odds ratios in Table 3) were grouped together and contrasted with whole milk consumption, whole full fat milk was found to be protective (OR = 0.67 95%CI 0.47-0.96: Table 4) and associated with good morphology. When red meat and whole full fat milk were entered into the same model, both retained their protective effect but this was reduced slightly for each factor individually as the two were correlated: those who ate red meat 3 or more times a week were also more likely to drink whole milk. The interaction between them was not significant in a fully adjusted model (last column, Table 4).
Consumption of soy, genistein and daidzein was not associated with poor morphology (Table 5).

TA B L E 3 Dietary factors for poor morphology
In this study, poor sperm morphology was associated with the lack of milk consumption and also the intake of skimmed or semiskimmed milk, suggesting that whole milk is protective of semen quality. Previously, a high intake of omega-3 polyunsaturated fats has been reported to be positively associated with sperm morphology 41 and fish oil supplements with semen volume and total sperm count. 42 Such results are consistent with the presence, in whole milk, of numerous fat-soluble vitamins and other components that might affect semen quality, for example, by providing protection against oxidative stress. 43 These would be lost upon processing to semi skimmed and skimmed milk. 44,45 Other studies, of smaller populations, have reported positive associations between consumption of low-fat milk and semen quality. 38,39 There are a number of strengths associated with this study.
Firstly, the number of men providing dietary information via a FFQ is much larger than many other previously published studies resulting in a study of sufficient power (80%) to detect an OR of 2 assuming 1 in 40 controls would be exposed, and a 2:1 ratio of referents to cases. The second strength of the study is that those men who knew the results of the assessment of their semen quality were excluded from the study. Such prior knowledge of their results could have biased their answers to specific questions or even modified behaviours prior to recruitment. Furthermore, semen analysis was carried out according to WHO protocols, with CASA being undertaken at each recruitment site but analysed centrally to ensure consistency in semen analysis. Case definitions were also as used previously 5,20 and defined a priori according to WHO  TA B L E 5 Soy and phytoestrogen intake with poor morphology men. In addition, not all eligible men were recruited to the study and the reasons why they did not want to participate are not known. It is possible that they did want to be asked about lifestyle factors (including diet). In addition to the recognized limitations of using FFQs to assess food consumption, which in this study was over a 12month period because of the sustained period of infertility required to be eligible for the study, there are potential issues regarding the questionnaire coverage of phytoestrogen containing foods and also the accuracy of the levels of phytoestrogens in the covered dietary items. 46 The food items and phytoestrogen values used here are not those of more recent studies (eg [47][48][49]. These issues are likely to be independent of case status and so result in non-differential bias, which is normally associated with a reduced estimate of effect size ('bias towards the null'). Assessment of phytoestrogen intake may be improved by urinary measurement of specific phytoestrogen metabolites; however, inconsistent results linking urinary daidzein and semen quality have also been reported. 50 54 It has also been suggested that pre-pubertal exposures may be important but we were not able to study them.
In conclusion, this large multi-centre case-control study identified different risk factors for low motile sperm count and poor sperm morphology. Further work is required to confirm these associations but suggests that semen quality may be improved by targeted interventions.

ACK N OWLED G EM ENTS
The study was designed and initiated with the support of Professor We are greatly indebted to the teams at each centre, which in-

CO N FLI C T O F I NTE R E S T
Professor JE Cade is a Director of the University of Leeds company, Dietary Assessment Limited.

AUTH O R CO NTR I B UTI O N S
All members of the co-ordinating group contributed to the collec-