The association between adult height and the incidence of diabetes in a Japanese working population: a prospective study exploring the possible role of cohort effects.

BACKGROUND

Adult height may reflect early-life environmental conditions that influence the risk of type 2 diabetes mellitus (DM). While taller height has been linked to lower DM risk, evidence from East Asia remains limited and inconsistent. Given that DM in East Asians often occurs at lower BMI levels, population-specific research is needed. We examined the prospective association between adult height and incident DM in a large Japanese working population, focusing on differences by sex and birth cohort.

METHODS

Data were obtained from annual health checkups of 63,264 workers from 12 companies in Japan, comprising 54,188 men and 9076 women aged 20-64 years at baseline. Participants were classified by quartile of adult height for each sex. DM was defined as either when the participants had glycated hemoglobin (HbA1c) ≥ 6.5%, random glucose ≥ 200 mg/dL, fasting glucose ≥ 126 mg/dL, or when they were receiving antidiabetic treatment. Time ratios (TRs) and 95% confidence intervals (CIs) for the duration of remaining free from DM were estimated using a sex-stratified accelerated failure time model. We also examined whether the association varied by birth cohort, by comparing individuals born in 1960 or earlier with those born later.

RESULTS

During a median follow-up period of 9.4 years, 5829 men and 448 women developed DM. For the highest vs. lowest height categories, after adjusting for age, body mass index, smoking, hypertension, and dyslipidemia, the TRs were 1.07 (95% CI, 1.02-1.13) in men and 0.98 (95% CI: 0.91-1.06) in women, respectively. When we stratified by birth cohorts, the height-DM association among men was more pronounced in the older cohort compared to the younger cohort.

CONCLUSIONS

In a large cohort of Japanese workers, taller stature was associated with a prolonged duration of DM-free survival among men, particularly those in older cohorts. In women, however, the association between height and diabetes risk was inconclusive and warrants further investigation. The impact of stature on the risk of DM may be affected by sex and cohort effects, necessitating further research to disentangle these effects.