ESPE Abstracts

Background: A secular trend for the timing of menarche has been described in women, but for men, studies of pubertal timing are scarce. Both negative and positive associations between childhood obesity and pubertal timing in men have been reported. In Sweden, Child Health Care (CHC) centers follow all children regarding growth and general health. We have collected detailed CHC growth data (height and weight) from centrally archived records for all children born 1946 or later in Gothenburg and established a unique population-based cohort, the BMI Epidemiology STudy (BEST; n≅400 000). The overall aim of the well-powered BEST cohort is to determine the role of childhood obesity and pubertal timing for a variety of diseases later in life.

Objective and hypotheses: The aim with the present BEST sub-study was to investigate if there is a secular trend for male pubertal timing and if childhood BMI influences pubertal timing.

Method: Men born every 5 years from 1946 to 1991 were evaluated (n=200 for each birth year, n=2000 for the entire sub-cohort). The height measurements were curve-fitted according to the Infancy – Childhood – Puberty (ICP) model and age at Peak Height Velocity (PHV) was calculated. PHV is the maximum growth velocity and represents a method to objectively determine age at pubertal timing.

Results: The mean age at PHV was 14.0±1.1 years. Linear regression analyses revealed that age at PHV was 1.3 months earlier for every 10 year increase in birth year between 1946 and 1991 (P=5.3×10⁻¹²). As expected, a secular trend of increased childhood BMI at 8 years of age was observed (P=1.2×10⁻¹²). We next evaluated the impact of childhood BMI on age at PHV and found that age at PHV was 2.1 months earlier for every quintile increase in BMI at 8 years of age (P=1.1×10⁻²¹). To determine the independent role of birth year and childhood BMI for age at PHV, both parameters were included in the same model, demonstrating that both birth year and childhood BMI were independently associated with age at PHV (standardized β; birth year β=−0.13, P=8.2×10⁻⁹; childhood BMI β=−0.21, P=9.8×10⁻²¹).

Conclusion: In conclusion, we provide compelling statistical evidence of a secular trend for pubertal timing in Swedish men. The secular trend of earlier pubertal timing is partly explained by increased childhood BMI but also other factors, to be identified, contribute.