Dietary quality in mid-pregnancy and pubertal timing in offspring: A population-based cohort study

Background and aims

Animal studies have suggested that nutrition received in utero may impact pubertal timing. However, the potential association between maternal dietary quality during pregnancy and pubertal timing remains unevaluated in humans. We examined if a measure reflecting dietary quality in mid-pregnancy is related to pubertal timing.

Methods

We utilized data from the Danish National Birth Cohort and its sub-cohort, the Puberty Cohort with children born in 2000–2003. Data on diet, the primary exposure, was collected in mid-pregnancy using a Food Frequency Questionnaire and condensed into eight domains based on Danish National Food-based Dietary Guidelines. Based on guideline compliance, a maternal healthy eating index (mHEI) ranging from 0 to 80 was constructed. Self-reported data on pubertal development, the primary outcome, was collected half-yearly from age 11 years and throughout puberty. Main outcome was age difference in months at reaching individual pubertal milestones as well as reaching all pubertal milestones in a combined estimate, which was analyzed according to mHEI in quartiles, continuous and as restricted cubic splines using multi-variable interval-censored regression models. The analyses were stratified by sex.

Results

The study population consisted of 12,979 mother-offspring pairs. The mHEI ranged from 4.5 to 54.7 with a median of 22.3. In girls, we found an indication of an association between mHEI and slightly earlier pubertal timing, as both high and low mHEI were associated with an earlier combined pubertal estimate compared to the median mHEI as restricted cubic splines. Not meeting fruit and vegetable recommendations was associated with later pubertal onset in girls. We found no association between mHEI and pubertal timing in boys.

Conclusion

There might be an association between mHEI and puberty timing in girls, but results should be interpreted cautiously due to the study's lack of consistent results, possible unmeasured residual confounding, a risk of misclassification and few participants meeting the dietary recommendations.