Pubertal timing and bone phenotype in early old age: findings from a British birth cohort study
Kuh D., Muthuri SG., Moore A., Cole TJ., Adams JE., Cooper C., Hardy R., Ward KA.
© The Author 2016. Published by Oxford University Press on behalf of the International Epidemiological Association. Objectives: To investigate the effect of pubertal timing, assessed in adolescence, on bone size, strength and density in men and women in early old age. Design: A British birth cohort study with prospective indicators of pubertal timing based on age at menarche, clinical assessment of pubertal stage, and growth tempo from serial height measures, and bone measures derived from peripheral quantitative computed tomography (pQCT) and dual-energy X-ray absorptiometry (DXA) at 60-64 years of age among 866 women and 792 men. Methods: A first set of regression models investigated the relationships between pubertal timing and bone size, strength and density, adjusting for current height and weight, smoking and adult socioeconomic position. To make an equivalent comparison between men and women, the percentage difference in bone outcomes was calculated for a 5-year difference in age at menarche, and in men a comparison between those who were fully mature or pre-adolescent at 14.5 years. A second set of models investigated the percentage difference in bone outcomes for a 5-year difference in timing of peak height velocity (height tempo) derived from longitudinal growth modelling (Superimposition by Translation and Rotation model; SITAR). Results: After adjustment for current height and weight, a 5-year increase in age at menarche was associated with an 8% [95% confidence interval (CI) -17%, 0.5%, P ¼ 0.07) lower trabecular volumetric bone mineral density (vBMD); men who were pre-adolescent at 14.5 years had a 9%, (95% CI -14%, -4%; P ¼ 0.001) lower trabecular vBMD compared with those who had been fully mature. Other confounders did not attenuate these estimates further. Patterns of association were similar but somewhat weaker for lumbar spine and total hip areal BMD. Age at peak height velocity was associated with even larger differences in BMD in men and women, and was negatively associated with bone size and strength. Conclusions: The association between later puberty and lower BMD persists into early old age. The 9-10% lower trabecular vBMD in later compared with earlier maturers could be clinically important given a rate of bone loss from midlife of 1-2% a year and the negative association between BMD and fracture.