Author Reply

Abstract

We sincerely appreciate Ren and Guo's thoughtful comments [1] on our study regarding prepregnancy healthy lifestyle and adverse pregnancy outcomes [2]. Their recognition of the study's significance, particularly in Asian populations, underscores the importance of addressing multiple lifestyle factors to improve maternal and infant health.

Assessment of prepregnancy lifestyle factors presents methodological challenges. Ideally, tracking women before conception would minimise bias, but this approach is constrained by the unpredictability of pregnancy onset and the time, budget and staffing demands of long-term follow-up. Given these limitations, the early pregnancy questionnaire survey was the most feasible option with the available resources. While we acknowledge that some women may have changed their behaviours upon recognising their pregnancy, potentially attenuating the observed association, retrospective assessments are widely performed in epidemiology, and the Japan Environment and Children's Study dataset provides a comprehensive evaluation of prepregnancy lifestyle factors. As Ren and Guo suggested, future studies incorporating prospective assessments and technologies for dynamic monitoring, such as wearable devices and biomarkers, could further enhance accuracy.

We agree that achieving an optimal healthy lifestyle (i.e., a Healthy Lifestyle Score [HLS] of 5) may not be feasible for all women due to socioeconomic and cultural constraints. However, the population attributable fraction (PAF) is a standard method to estimate the potential public health impact under idealised conditions. While we adjusted for educational attainment as a proxy for socioeconomic status (SES), residual confounding remained, highlighting the need for more comprehensive SES evaluation. Our estimated PAF aligns with the previous study using the Nurses' Health Study II, which reported a PAF of 13% based on six low-risk prepregnancy healthy lifestyle factors [3]. Additionally, we calculated the PAF₊₁, estimating the preventable fraction assuming all but women with an HLS of 5 achieved a one-point increase, which offers clearer insights into the potential benefits of incremental lifestyle improvements.

Although non-linearity and potential threshold effects cannot be excluded, the observed risk reduction from HLS of 2 and the consistent dose–response relationship in our study [2] and the prior research in the United States [3] suggest that a linear model sufficiently captures this association. Future studies should explore these possibilities further, particularly in relation to socioeconomic and psychosocial stratification, to refine theoretical understanding and inform targeted interventions.

Ren and Guo's [1] valuable suggestions on public health interventions, particularly the role of social workers, are critical to translating research into practice. While our study primarily focused on epidemiological associations, future research should explore strategies to promote behaviour change, especially among those with unhealthy lifestyle or facing socioeconomic challenges. A multi-level approach, incorporating individual education, family and community support as well as policy-level initiatives, will be crucial to improving prepregnancy health outcomes [4]. Furthermore, consideration of variations in lifestyle sensitivities based on SES could help identify more effective interventions, with social workers playing a key role.

H.O. wrote the author's reply. The other authors reviewed and approved the final manuscript.

The authors declare no conflicts of interest.