Does Shrimp Really Boost Your Sexual Drive? Here’s What Science Says!

Audrey J Gaskins1Department of Nutrition, Harvard T.H. Chan School of Public Health, Boston,Massachusetts2Channing Division of Network Medicine, Department of Medicine, Brigham andWomen’s Hospital and Harvard Medical School, Boston, MassachusettsFind articles by

Received 2018 Feb 16; Accepted 2018 May 1; Collection date 2018 Jul. Copyright © 2018 Endocrine Society

Marine long-chain omega-3 fatty acids have been positively related to markers of fecundity in both men and women. However, seafood, their primary food source, can also be a source of toxicants, which could counteract the reproductive benefits.

To examine the relationship of male and female seafood intake with time to pregnancy (TTP).

Our prospective cohort study included 501 couples planning pregnancy, who participated in the Longitudinal Investigation of Fertility and the Environment study (2005 to 2009) and were followed up for ≤1 year or until pregnancy was detected. Seafood intake was collected daily during follow-up in journals.

Couples residing in Michigan and Texas were recruited using population-based sampling frameworks.

The primary outcome was the TTP, determined using an in-home pregnancy test. A secondary outcome was sexual intercourse frequency (SIF) as recorded in the daily journals.

Couples with male and female partners who consumed eight or more seafood servings per cycle had 47% (95% CI, 7% to 103%) and 60% (95% CI, 15% to 122%) greater fecundity (shorter TTP) than couples with male and female partners who consumed one or fewer seafood servings per cycle. Couples with both partners consuming eight or more seafood servings per cycle had 61% (95% CI, 17% to 122%) greater fecundity than couples consuming less. Male and female partners with the highest seafood intake (eight or more servings per cycle) also had 22% greater SIF.

Greater male and female seafood intake was associated with a higher SIF and fecundity among a large prospective cohort of couples attempting pregnancy.

In the present prospective cohort study with preconception enrollment and daily follow-up of couples, seafood intake in both partners was associated with a higher SIF and greater fecundity.

Infertility, the failure to achieve pregnancy after 12 months of unprotected sexual intercourse, affects 15% to 25% of couples (1, 2). Although infertility treatments exist, their costs (3), limited geographic accessibility (4), and modest success (5) justify identifying modifiable factors that increase a couple’s chance of conceiving without medical assistance. Seafood is a recommended component of many healthy eating patterns (6, 7). In the context of fertility, however, seafood has largely been studied as a potential harm, representing a primary source of exposure to reproductive toxicants such as organochlorines, dioxins, and mercury (8–11). In contrast, some studies have found reproductive benefits with higher marine long-chain omega-3 fatty acid intake, such as increased progesterone levels, a shorter time to pregnancy (TTP), and better semen quality (12–14).

For the average US adult, the current recommendation is to eat at least two seafood servings per week (6); however, in January 2017, the US Food and Drug Administration and Environmental Protection Agency recommended that women who are pregnant or might become pregnant should eat no more than three servings per week (15). This guideline was established to limit fetal methyl-mercury exposure, which has been linked to adverse neurocognitive consequences. However, to the best of our knowledge, these guidelines did not consider the potential reproductive benefits of seafood intake. To address this gap, we used data from a prospective cohort of couples attempting to become pregnant, with information on daily seafood intake and sexual intercourse collected in journals, to investigate whether male and female seafood intake was associated with the TTP and whether this association could be due to differences in sexual activity.

The Longitudinal Investigation of Fertility and the Environment (LIFE) study is a prospective cohort of 501 couples attempting to conceive in two geographic areas (Texas and Michigan) from 2005 to 2009. Using population-based sampling frameworks, households were contacted to identify eligible couples in a committed relationship. Female partners were required to be aged 18 to 44 years, to have menstrual cycles of 21 to 42 days, and to have had no hormonal birth control injections during the previous year. Male partners were required to be aged ≥18 years. Both partners were required to have the ability to communicate in English or Spanish and to have undergone no sterilization procedures or have physician-diagnosed infertility. The couples were also excluded if they had not been using contraception for >2 months. A complete description of the study’s methods has been previously reported (16). In brief, of the 1188 eligible couples, 501 (42%) were enrolled in the present study and followed up for ≤12 months, with monthly pregnancy tests. The institutional review boards at each institution approved the protocol. All participants provided written informed consent.

Research assistants traveled to the couples’ homes and completed baseline in-person interviews separately with each partner. Both partners were asked how often during the previous 12 months they had eaten canned tuna fish; fish caught in unknown locations; crab, shrimp, or other shellfish caught in an unknown location; fish caught in local waters; and crab, shrimp, or shellfish caught in local waters. The five response options ranged from “never or almost never” to “two or more times per week.” The selected frequency category for each seafood item was then converted to a monthly intake, and all items were summed to find the total baseline seafood intake. In the daily journals, the male and female participants were asked to report the number of 4-oz servings of fish or shellfish consumed. These daily responses were then summed across the cycle to determine their cycle-specific seafood intake. For the analysis, the baseline and daily male and female seafood intake were classified into categories that approximated quartiles.

During the enrollment interview, each partner reported their age, level of education, ethnicity, race, household income, and use of cigarettes. Participants were asked whether they had followed a regular vigorous exercise program in the previous 12 months and, if so, how many days per week. The four-item Cohen perceived stress scale was also administered (17). The men and women reported whether they had consumed ≥12 alcoholic drinks in the previous year, and, if so, how often they had consumed alcoholic beverages. All participants had their weight and height measured using standardized procedures, and the body mass index was calculated as the weight in kilograms divided by the height in square meters.

The primary outcome was fecundity, as measured by TTP. We used daily journal information supplemented with fertility monitors to define the menstrual cycles, defined as the interval (in days) from the onset of bleeding that increased in intensity and lasted ≥2 days to the onset of the next similar bleeding episode. Because couples were allowed to enroll in the LIFE study midcycle, we defined this as cycle 0 to differentiate it from cycle 1, which denoted the first fully observed menstrual cycle. Pregnancy was defined as a positive study-provided home pregnancy test, which was sensitive for 25 mIU/mL human chorionic gonadotropin. A secondary outcome was the frequency of vaginal–penial intercourse, as recorded by the men and women in their daily journals. For each cycle of follow-up, the sexual intercourse frequency (SIF) reports were summed across all days to find the total SIF per cycle. The correlation between the SIF per cycle as reported by the male and female partners was 0.98, and the average difference between the two reports was −0.02 times per month. Because of the slightly lower amount of missing data in the female diaries, the female report of SIF was used as the main outcome variable.

We classified each partner as having high (nine times or more per month; 75th percentile) or low-to-average (less than nine times per month) seafood intake. The male and female demographic data and lifestyle characteristics were then compared using ANOVA for continuous variables or χ2 tests for categorical variables. The correlation within and between male and female seafood intake at baseline and during follow-up was calculated using Spearman correlation coefficients.

Cox proportional odds models for discrete survival data accounting for left truncation (to account for the time without contraception before enrollment) and right censoring (to account for the loss to follow-up or the end of the study) were used to estimate the fecundability ORs (FORs), and their 95% CIs, as a measure of fecundity. FORs represent the relative odds of achieving pregnancy conditional on not becoming pregnant in the previous cycle, such that an FOR <1 indicates diminished fecundity as measured by a longer TTP. Seafood intake was initially considered as quartiles of intake, and in a supplemental analysis, it was modeled continuously using linear and quadratic terms.

To analyze the association between seafood intake and SIF per cycle during the follow-up period, we used generalized linear mixed models with the Poisson distribution. Effect estimates and 95% CIs are presented as the percentage of difference in SIF for a particular group compared with the reference group. We also explored the association between day-level seafood intake and SIF using a generalized linear mixed model with logit link. The results are presented as the ORs and 95% CIs of sexual intercourse in a given day. We imputed the SIF values for cycles with >50% of the days missing information on SIF and any cycle with <14 days of follow-up (n = 159 cycles) using Markov chain Monte Carlo methods (PROC MI in SAS; SAS Institute, Cary, NC) with five multiple imputations based on menstrual cycle length, cycle number of follow-up, female age, the difference between the couple’s ages, female race and education level, and male exercise. Effect estimates from models using multiply imputed values for SIF were estimated using Rubin’s formula for combining estimates across imputations (PROC MIANALYZE in SAS).

Confounding was evaluated using previous knowledge and descriptive statistics from our cohort through the use of directed acyclic graphs. Variables retained in the final multivariable models were female age (in years), the difference between couple’s ages (in years), female race (non-Hispanic white vs other), male exercise (yes vs no), and male and female alcohol intake (one or more time per week vs less than one time per week). Additional models were run further, adjusting for male and female partner seafood intake owing to the high amount of concordance within a couple. The fecundity models were also further adjusted for SIF to evaluate the extent to which this variable explained any observed associations. A P value for trend was calculated across the categories of seafood intake using the median intake level in each category as a continuous variable.

In the main analysis, missing data on seafood intake in the daily journals were considered as no intake, which is common for dietary analyses. Sensitivity analyses were performed in which missing seafood intake was imputed for cycles that were missing 100% and >50% of days of seafood intake data using Markov chain Monte Carlo methods with five multiple imputations and Rubin formula to combine estimates across imputations. To address concerns of residual confounding, we also calculated propensity scores and ran the final model adjusting for this variable and stratified by quintiles of this variable. To quantify the effect of unmeasured confounding, we calculated the e-value, which estimates the minimum strength of an association that an unmeasured confounder would need to have with both the exposure and outcome to fully explain a specific exposure–outcome association (18). SAS, version 9.4 (SAS Institute) was used for all statistical analyses.

Male partners who reported the highest usual seafood intake were less likely to have a non-Hispanic white partner and more likely to exercise regularly and consume alcohol one or more time per week compared with men with lower intake (Table 1). Female partners with the greatest usual seafood intake were, on average, older, had older partners, were less likely to be non-Hispanic white, and were more likely to consume alcohol one or more time per week compared with females with lower intake. Seafood intake was not associated with body mass index, education level of either partner, or household income. Male and female seafood intake within a couple correlated moderately at baseline (r = 0.46) and correlated highly during the follow-up period (r = 0.70; Supplemental Table 1). Within men and women, the baseline seafood intake correlated moderately with the intake during follow-up (r = 0.47 and r = 0.53, respectively).

Demographic and Lifestyle Characteristics Stratified by Seafood Intake at Baseline in the LIFE Study (n = 501 Couples)

Higher male (but not female) baseline seafood intake was associated with higher SIF during follow-up after multivariable adjustment (Table 2). Men who usually consumed seafood nine or more times per month had a 22.9% (95% CI, 6.8% to 41.5%) greater SIF compared with men who usually consumed seafood two times or less per month (P for trend = 0.007). A positive association was found between baseline female seafood intake and SIF that became attenuated after adjustment for male partner intake. During follow-up, both male and female seafood intake was independently associated with SIF, with slightly stronger associations observed for male intake. Furthermore, when both partners consumed eight or more servings per cycle, SIF was increased by 21.9% (95% CI, 15.2% to 29.0%) compared with couples consuming less. In the day-level analyses, the odds of sexual intercourse was 39% (95% CI, 29% to 50%) greater if both partners consumed seafood the same day, 3% (95% CI, −5% to 11%) greater if only the woman consumed seafood, and 2% (95% CI, −6% to 10%) greater if only the man consumed seafood compared with couples with neither partner consuming seafood. The associations were identical when the male report of SIF was used (instead of the female report).

Associations Between Male and Female Seafood Intake at Baseline and During Follow-Up and Frequency of Sexual Intercourse (n = 501 Couples; 2372 Follow-Up Cycles)