Sedentary behavior, physical inactivity and body composition in relation to idiopathic infertility among men and women.

Sedentary behavior and physical inactivity represent major health concerns [ 1 – 3 ]. Sedentary behaviors are defined as any waking activities characterized by energy expenditure below 1.5 metabolic equivalent of task (MET) of sitting or lying down [ 4 ]. Physical inactivity represents an insufficient volume of physical activity (PA) in daily life, being a level not reaching the recommended PA (150 minutes of moderate PA per week) [ 4 ]. These two behaviors are in some cases coexistent, and sometimes not. Indeed, an individual may have both sedentary behaviors and be physically active [ 5 , 6 ]. In this case, PA can moderate but not offset the deleterious effects of sedentary behavior [ 1 ]. It has been shown that sedentary behaviors and physical inactivity independently influence several health factors, non-communicable diseases and mortality [ 1 – 3 , 7 , 8 ]. Notably, PA has an inconsistent effect on fertility. In men, moderate PA has been positively associated with semen quality [ 9 – 12 , 13 ]. However, it was not associated with higher reproductive success in the context of fertility treatment [ 12 ]. Some previous studies failed to demonstrate a relationship between PA and semen quality [ 14 , 15 ]. In women, moderate PA increased fecundity parameters and live birth rates, regardless of body mass index (BMI) [ 16 , 17 ]—even during assisted reproductive treatment [ 18 – 20 ]. However, vigorous activity has been associated with lower semen quality in men [ 21 – 23 ] and decreased fertility in women [ 24 – 26 ]. Notably, sedentary behavior has not been clearly associated with semen quality [ 12 , 13 , 20 , 21 , 27 , 28 ], though reduced sperm concentration has been linked to increased television watching [ 11 ]. In women, sedentary behavior has not been associated with lower fertility in recent studies [ 20 , 29 ]. Obesity is associated with both sedentary behavior and physical inactivity [ 30 , 31 ]. Being overweight and obese is known to impact the fertility of couples [ 32 ]. Large cohort studies showed that a BMI over 25 kg/m 2 (as estimated by the height/weight 2 ratio) was linked to infertility in both males and females [ 32 – 34 ]. Obesity has been associated with reduced semen quality [ 35 ], sperm concentration [ 33 , 36 – 38 ], mobility [ 39 ], DNA damage [ 40 – 42 ], poor oocyte quality, and impaired ovulation and implantation [ 34 ]. In the aforementioned studies, obesity estimation was based on BMI values. However, anthropometrics are not the most sensitive parameters for estimating body composition alterations [ 43 ]. To our knowledge, very few studies have explored body composition or adiposity in association with fertility, especially fat mass and fat-free mass parameters. Recent studies have used waist circumference and BMI as proxy measures of body composition [ 44 , 45 ]; one used dual-energy X-ray absorptiometry for fat, fat-free mass, and bone mass in 41 young infertile women [ 46 ]. Idiopathic or unexplained infertility is defined when a lack of diagnosis appears in couples that failed to conceive after 1 or 2 years of non-protected sexual intercourses [ 47 ]. It concerns 30 to 40% of infertile couples [ 48 ]. Standard investigation protocol of idiopathic infertility involved tests of ovulation, tubal patency and semen analysis. The diagnosis of idiopathic infertility may be very frustrating for the couples and the treatment is usually empirical [ 47 ]. Even if no cause is clearly identified, the environment and lifestyle could be an explanation of some idiopathic infertility. Consequently, it is important to focus on modifiable risk factors in this population. Better understanding the origin of the disorder should be a way to manage idiopathic infertility. The primary objective of this study was to determine if physical inactivity, sedentary behavior and body composition were related to idiopathic infertility in men and women in a French case-control study of nutritional determinants of idiopathic infertility.

Baseline characteristics of the 302 participants are presented in Table 1 . Infertile participants were younger in comparison to fertile men and women ( p = 0.006 and p = 0.02, respectively). They also had lower educational levels than fertile men and women ( p = 0.005 in men and p = 0.0001 in women). The weight, BMI, waist circumference, hip circumference, and body fat of infertile men and women were significantly higher compared to fertile men and women. In men, the proportion of participants with metabolic syndrome was higher in infertile compared to fertile participants [12 (16.0%) vs . 3 (4.4%), respectively p = 0.03]. The proportion of normal weight obese did not differ between groups of fertile and infertile men [5 (6.9%) and 5 (6.3%), respectively, p = 1] and women [22 (31.0%) and 20 (25.0%), respectively, p = 0.5)]. Mean PA levels did not significantly differ between fertile and infertile men (2726.2 and 3291.2 MET-min/week, respectively, p = 0.7) and women (2632.8 and 2769.4 MET-min/week, respectively, p = 0.8). However, infertile men spent less time performing vigorous PA (37.6±48.6 min/week vs . 69.3±84.4 min/week, p = 0.006) in comparison to fertile men. Mean walking time was not different in infertile men compared to fertile men (42.3±73.8 min/week vs . 35.9±36.4 min/week, p = 0.2), nor in infertile women compared to fertile women (29.7±34.4 min/week vs . 46.6±66.5 min/week, p = 0.08). Physical activity was only inversely associated with sedentary behavior in infertile men (r Pearson = -0.3, p = 0.04). Physical activity was only inversely associated with body fat percentage in fertile men (r Pearson = -0.3, p = 0.03). All infertile and fertile participants followed nutritional guidelines similarly, with scores of 6.6±2.1 vs. 6.2±1.9, respectively, for men ( p = 0.4), and scores of 6.3±2.9 vs. 6.3±3.1, respectively, for women ( p = 0.9) (maximal possible score of 15). Based on PA guidelines, 34 (47.2%) and 50 (63.3%) ( p = 0.05) fertile and infertile men did not follow PA guidelines (150 min/week of moderate-to-vigorous PA), respectively. Moreover, 43 (60.6%) and 55 (68.8%) ( p = 0.3) fertile and infertile women were under the recommended PA level, respectively. a Independant t test b Wilcoxon-Mann-Whitney test c Fisher exact test d BMI ≥ 25 kg/m 2 e BMI < 25 kg/m 2 but fat mass over reference values for age and gender f 80 cm for women, 94 cm for men PA, sedentary behavior, and body composition factors according to fertility status and gender are presented in Table 2 . In men, being physically inactive (adjusted OR 2.20; 95% CI, 1.06, 4.58; p = 0.04) and having excess body fat (adjusted OR 2.83; 95% CI, 1.31, 6.10; p = 0.008) were positively associated with infertility. Sedentary behavior and fat-free mass were not related to infertility in men in our study. In women, having sedentary behavior (adjusted OR 3.61; 95% CI, 1.58, 8.24; p = 0.002) and having body fat over (adjusted OR 3.16; 95% CI, 1.36, 7.37; p = 0.008) and fat-free mass under (adjusted OR 2.65; 95% CI, 1.10, 6.37; p = 0.03) reference values for their age were associated with a significantly increased risk of infertility. Physical activity was not significantly associated with fertility status among women in our study ( p = 0.3). Abbreviations: OR, Odds ratio, Adj OR, Adjusted Odds ratio, CI, Confidence Interval, PA, Physical Activity, SD, Standard Deviation. a Adjusted for age and educational level and for all variables of the table. b Age and gender reference values [ 53 ].

The present study demonstrated that physical inactivity in men and sedentary behavior in women are associated with idiopathic infertility. Body fat accumulation has been related to infertility in both men and women, while fat-free mass was related to infertility in women only. This case-controlled study highlights that physical inactivity and sedentary behavior represent two independent risk factors for infertility. The effect of various elements that make up PA (i.e. FITT criteria) and interrupting the time spent sitting were not tested in this study, and should be considered in future research. The differences observed between men and women should also be studied further through multicentric interventional studies to better understand the lifestyle to promote to men and to women respectively. Moreover, body composition variation through lifestyle should be also explored further in relation to the biological pathways involved in idiopathic infertility. These findings suggest promoting and proposing a lifestyle supportive care during fertility treatment in order to improve pregnancy rates.

Participants were recruited in the ALIFERT case-control multicentric observational study (“ALImentation et FERtilité”, ClinicalTrials.gov identifier: NCT01093378 ), which evaluated the associations between nutritional parameters and fertility among infertile and fertile couples. The institutional review board approved the study (ALIFERT study—national biomedical research P071224/AOM 08180: NEudra CT 2009-A00256-51). Data were recorded from 302 French participants, with included 159 infertile (79 men and 80 women) and 143 fertile (72 men and 71 women). Men under 45 years of age and women under 38 years of age were included. Infertile participants had a history of primary idiopathic infertility for at least 12 months of unprotected sexual intercourse, with no diagnosed etiology for their infertility. They never had history of miscarriages and did not start infertility treatment at inclusion. Men were excluded if they had severe oligozoospermia (< 5 million/mL), azoospermia, or any abnormality of the male genital tract (undescended testis or varicocele). Women were excluded if they presented anovulation, ovarian failure, uterotubal pathology, or endometriosis. Fertile participants had a recent natural and spontaneous pregnancy and delivery (< 24 months) with a time to conceive shorter than 12 months. No specific matching was conducted between cases and controls, and one fertile control couple was selected for each case couple. Participants completed self-administered questionnaires on sociodemographic and lifestyle characteristics (sex, age, educational level, and smoking status), dietary intake (semi-quantitative validated food frequency questionnaire), physical activity and sedentary behavior. Anthropometrics, body composition, and blood pressure were measured using standardized procedures (tensiometer; Omron M5-I). Blood samples were used to evaluate metabolic syndrome with plasma high-density lipoprotein (HDL), triglycerides and fasting glycaemia in mmol/L. Assessments were performed after an 8-hour fasting period. PA level and sedentary behavior were estimated by the self-administered validated last-7-day International Physical Activity Questionnaire (IPAQ) [ 49 ]. PA levels correspond to the PA level of a typical week during the inclusion period. Total PA level scores were expressed in metabolic equivalent of task (MET) per minute per week (MET-min/week), calculated as a duration X frequency per week X MET intensity of PA retrieved from the items of the IPAQ questionnaire for moderate PA, vigorous PA (occupational and leisure time) and walking activities (in min/week). Accordance with guideline targets (150 min/week of moderate-to-vigorous PA) was estimated by adding times of moderate, vigorous and walking activities (in min/week). Sedentary behavior was assessed through a question regarding time spent sitting during typical week days (in h/day). A threshold of 5h per day was chosen to categorize participants as having sedentary behavior (≥ 5h/day) or not (< 5h/day). This threshold corresponds to the average time spent while sitting (when occupational time is included) in the general French population [ 50 ]. The height and weight of participants was measured to the nearest 0.5 cm and 0.5 kg, respectively, with participants wearing light clothing and no shoes using standardized procedures. BMI (kg/m 2 ) was calculated as the weight (kg) divided by the square of height (m). Patients with a BMI over or equal to 25 kg/m 2 were considered as overweight. Waist and hip circumferences were measured using a measuring tape accurate to 0.1 cm. Measurements were performed by a trained investigator during the morning under fasting conditions. Body composition was estimated by single frequency bioelectrical impedance analysis 50 kHz (Tanita BC 420 S MA, Tanita Corp., Tokyo, Japan). It combines a digital scale with stainless steel pressure-contact footpad electrodes for standing impedance and body weight measurements [ 51 ]. The measurement relies on the differences in resistance after the conduction of an electrical current through the body. It enables a rapid assessment of body composition without radiation [ 52 ]. Body fat percentage (%) and fat-free mass (kg) were assessed. Reference values of body fat percentage and fat-free mass in healthy European subjects [ 53 ] were used to estimate if individuals had excess body fat and a lack of fat-free mass according to their age and sex. Participants with excess body fat despite exhibiting a normal BMI (< 25 kg/m 2 ) were considered as “normal weight obese” [ 54 ]. The validated Programme National Nutrition Santé Guideline Score (PNNS-GS) was used to consider the adherence of individuals to the French dietary guidelines for fruits and vegetables, starchy foods, milk and dairy, meat, fats, sweetened foods, beverages, salt intake, and PA [ 55 ]. The maximum score was 15. If a participant possessed three or more of the following risk factors, they were considered to have metabolic syndrome according to the International Diabetes Federation (IDF) and American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) thresholds [ 56 ]. Risk factors included: a waist circumference ≥ 94 cm in men and ≥ 80 cm in women; low HDL < 1.03 mmol/L in men and < 1.29 mmol/L in women; elevated triglycerides ≥ 1.7 mmol/L; elevated fasting glycemia ≥ 5.6 mmol/L; and elevated blood pressure (systolic blood pressure ≥ 130 mmHg and diastolic blood pressure ≥ 85 mmHg). The baseline characteristics of participants were described by gender and fertility status (frequency and percentage of categorical variables, as well as the mean and standard deviation of quantitative variables). Men and women were analysed separately due to their differences in lifestyle and body composition, as well as the different physiological benefits of exercise in both genders [ 57 – 59 ]. Comparisons between case and controls were conducted using Fisher’s exact test (as appropriate for categorical variables) and independent t -test or Wilcoxon-Mann-Whitney test when appropriate (for continuous variables). Pearson correlation coefficients were computed to assess the relationship between PA level and sedentary behavior, and between PA level and body fat percentage. Analyses were performed separately for men and women. Associations between PA and sedentary behaviour with fertility status were investigated using logistic regression models. Multivariable analyses were performed after crude univariate logistic regressions. We elected not to include more than six covariates in the final model (age, education level, PA level, sedentary behavior, body fat and fat-free mass) in accordance with the literature [ 57 ] and to adhere to the principle of one variable studied for ten cases in small sample study. The regression model was adjusted for these six variables. Unadjusted and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were reported. BMI and waist circumference was not included in the models due to collinearity with body composition. SAS version 9.1 (SAS institute, Cary, NC, USA) was used to perform for all statistical analyses. A p <0.05 was considered significant.