Are There Indicators of Risk for Adenomyosis Already in Early Life?

Adenomyosis was first described in the mid-19th century and recognised as a separate disease entity in the mid-20th century. Nonetheless, its aetiology remains poorly understood. The disease is characterised by the presence of endometrial glands and stroma within the uterine myometrium and is most commonly associated with abnormal uterine bleeding, chronic pelvic pain and infertility. It additionally adversely impacts women's quality of life [1]. Estimates of the burden of adenomyosis vary widely across populations and depend on the diagnostic criteria used; it is substantially underdiagnosed. With estimates ranging from 1% to 70% of women, the true global prevalence is unknown [2]. In this issue of Paediatric and Perinatal Epidemiology, Hall et al. [3] investigate whether a wide range of potential intrauterine and early-life factors are associated with adenomyosis. Underpinning this investigation is a theoretical link between in utero influences on oestrogen production (through effects on the hypothalamus-pituitary-ovarian axis), contributing to adenomyosis's pathogenesis. The authors conducted a case–control study using a population of women from Washington State, USA, aged 18–59 years, at enrolment into the Keiser Permanente Healthcare System. They used a group of 386 women with adenomyosis and compared it with a group of 233 women who underwent a hysterectomy and a group of 323 women from the healthcare population. The authors examined nine early-life exposures, ranging from the mother's age and smoking status to the participant's birthweight and birth order. Due to statistical instability, however, the authors did not report associations for all of the exposures they initially intended to. Among the intriguing findings were indications of increased odds of adenomyosis with younger maternal age at birth, later birth order and maternal smoking (only among never-smoking participants). Additionally, the authors found some indications of positive associations with birthweight. However, all risk estimates had wide confidence intervals, suggesting imprecision. Two previous Scandinavian cohort studies found limited evidence for associations between early-life risk factors, including birthweight, and risks of adenomyosis [4, 5]. Even if birthweight was not associated with adenomyosis risks in the previous investigations, a study conducted within our research group found that body mass index in childhood may be a risk factor for subsequent risks of adenomyosis [5]. Hence, this emphasises that examinations of early-life factors, exceeding beyond intrauterine and perinatal life, also warrant consideration as indicators of adenomyosis risk. Among the strengths of the study by Hall et al. is the inclusion of pathology-confirmed adenomyosis by hysterectomy, which is considered the gold standard for diagnosing adenomyosis. However, as acknowledged by the authors, there is also a risk of women having undiagnosed adenomyosis among population controls, as imaging technologies were not available when the study was conducted. Another strength is that information about a range of potential confounding factors was available, and the authors performed several sensitivity analyses. Some important limitations of the case–control study by Hall et al. should be considered. The exposure information was from self-reported recall of participants (or in combination with their mothers), which may introduce potential exposure misclassification. Although early-life factors may contribute to the subsequent development of adenomyosis, the findings by Hall et al. remain open to interpretation as the confidence intervals were wide and included many possible risk values. For example, the adjusted odds ratio of later versus first-born child was 1.51 (95% confidence interval 0.88, 2.59), which is compatible with an increased odds of 51%. Although this is the most likely odds, the imprecise 95% CI suggests a wide margin of the potential odds: from 12% reduced odds (OR: 0.88) to 159% increased odds (OR: 2.59). Due to the small number of cases, the risk estimates have relatively high uncertainty, which precludes firm conclusions. Therefore, future studies should include populations of sufficient size with available and accurate information on both exposures and outcomes to be able to interpret if and how early-life factors may relate to the subsequent development of adenomyosis, preferably using cohort designs. Adenomyosis is a debilitating disease with a substantial impact on women's lives and further constitutes a high burden to health care systems. Unfortunately, this disease is understudied compared with many other benign gynaecological reproductive conditions. With advances in imaging technology, non-invasive diagnostics are now possible. As diagnoses of adenomyosis are no longer reliant on uterine specimens from hysterectomies, it is emerging that adenomyosis may occur earlier in a woman's life than previously thought. The study by Hall et al. is timely and highlights the relevance of early-life and even intergenerational investigations into the origins of adenomyosis. By identifying risk factors early in life in combination with the possibility of detecting adenomyosis at younger ages, better health outcomes for women afflicted with this debilitating disease may be ensured. The author thanks Jennifer L. Baker for her comments on an earlier draft of this commentary. The author declares no conflicts of interest. The author has nothing to report.