Inserm, U823, IAB, Grenoble, France, Institut Albert Bonniot, Grenoble University, Grenoble, France (Philippat, Giorgis-Allemand)
Inserm, U625, GERHM, Rennes, France, GERHM, IFR140, University of Rennes I, Rennes, France (Chevrier, Cordier, Jégou)
Inserm, U1018, CESP, Villejuif, France (Charles)
Inserm, U823, IAB, Grenoble, France, Institut Albert Bonniot, Grenoble University, Grenoble, France, Remy.email@example.com (Slama)
To the Editor:
Two recent papers suggested associations between the use of analgesics (paracetamol, aspirin, and ibuprofen) by women during pregnancy and the occurrence of congenital cryptorchidism in their offspring.1,2 This hypothesis deserves consideration given the widespread use of analgesics by pregnant women and the suspected secular increase in incidence of undescended testes in some areas.3 We therefore explored this association within a cohort of pregnant women in which specific information on the position of the testis has been collected at birth.
Women from Eden mother–child cohort were recruited before 24 gestational weeks in the maternity wards of Nancy and Poitiers (France) University hospitals, between 2003 and 2006.4,5 Exposure to analgesics during the first 2 trimesters of pregnancy was assessed from a questionnaire asked by the study midwives between 24 and 28 gestational weeks. Women were asked “Since the beginning of the pregnancy, did you take one of the following drugs? Aspirin, paracetamol (listing 3 common preparations containing paracetamol) or ibuprofen (Advil).” Answers were recorded as yes/no/does not know, without distinguishing among the types of analgesic. A similar question was asked by the study midwife shortly after birth regarding analgesic use during the whole pregnancy. During a clinical examination conducted within 3 days after birth, pediatricians and study midwives assessed testis location (in scrotum, inguinal superficial, or not palpable). All singleton newborns with at least one testis not located in the scrotum were categorized as cases of undescended testes.
We studied the association between analgesics use and occurrence of undescended testes, using logistic regression adjusted for gestational duration (defined from the date of the last menstrual period5,6), center, maternal age, parity, smoking, and educational level. Propensity-based methods provide an alternative to traditional adjustment to correct for confounding.7 Such methods may be particularly relevant when the number of cases (and hence the maximum number of covariates one can efficiently control for by adjustment8) is limited. We built a propensity score corresponding to the probability of use of analgesics as a function of maternal characteristics, and used this in additional weighted logistic models quantifying the impact of analgesics on occurrence of undescended testes using the inverse-probability-of-treatment weighted estimator.7
The frequency of undescended testis was 3.4% in Poitiers (16 cases, 465 births) and 5.1% in Nancy (22 cases, 430 births), giving an overall rate of 4.3% (4.1% after exclusion of preterm births). Analgesic use (overall frequency, 81%) was associated with parity (P = 0.04), toothache (P = 0.02), and antibiotic use (P = 0.08) during pregnancy. Frequency of undescended testis was 4.6% among offspring exposed to analgesics, compared with 2.9% among unexposed offspring (unadjusted odds ratio [OR] = 1.6 [95% confidence interval (CI) = 0.6–5.3]). After adjustment, the OR of undescended testis associated with use of analgesics during the 2 first trimesters of pregnancy was 1.2 (CI = 0.5–3.2; Table) in the whole population, and 1.5 among term births (CI = 0.5–4.4). ORs associated with analgesic use during the whole pregnancy were in the same range. The propensity-based weighting approach yielded ORs close to 1.0 (Table).
The main strength of our study is its longitudinal design. Limitations include small sample size, an inability to distinguish among specific compounds, and lack of information on dose, mixture, and exact timing of use. Our estimate of the association between maternal use of analgesic during the whole pregnancy and undescended testis risk (OR = 1.5 [95% CI = 0.5–7.5]; Table) was similar to that reported in the Danish population described by Kristensen et al1 before adjustment (OR = 1.5 [95% CI = 0.8–2.8]), but weaker after adjustment (OR = 1.2 in our study vs. 1.4 in the Danish study). We adjusted for parity, as was done by Jensen et al2 but not Kristensen et al. Our point estimates seemed sensitive to the choice of adjustment factors, to the exclusion of preterm births and propensity score weighting. Confidence intervals were broad, not allowing for conclusions regarding a possible impact of analgesics on the risk of undescended testes. We suggest that other cohorts with relevant exposure and outcome data report similar analyses, so that a meta-analysis can be performed.
Inserm, U823, IAB
Institut Albert Bonniot
Inserm, U625, GERHM
University of Rennes I
Inserm, U1018, CESP
Inserm, U823, IAB
Institut Albert Bonniot
1.Kristensen DM, Hass U, Lesne L, et al. Intrauterine exposure to mild analgesics is a risk factor for development of male reproductive disorders in human and rat. Hum Reprod
2.Jensen MS, Rebordosa C, Thulstrup AM, et al. Maternal use of acetaminophen, ibuprofen, and acetylsalicylic acid during pregnancy and risk of cryptorchidism. Epidemiology
3.Boisen KA, Kaleva M, Main KM, et al. Difference in prevalence of congenital cryptorchidism in infants between two Nordic countries. Lancet
4.Drouillet-Pinard P, Huel G, Slama R, et al. Prenatal mercury contamination: relationship with maternal seafood consumption during pregnancy and fetal growth in the ‘EDEN mother-child’ cohort. Br J Nutr
5.Slama R, Thiebaugeorges O, Goua V, et al. Maternal personal exposure to airborne benzene and intrauterine growth. Environ Health Perspect
6.Slama R, Khoshnood B, Kaminski M. How to control for gestational age in studies of effects of environmental factors on fetal growth? Environ Health Perspect
7.Kurth T, Walker AM, Glynn RJ, et al. Results of multivariable logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect. Am J Epidemiol
8.Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and cox regression. Am J Epidemiol