Abstract

Small vertebral size is an independent risk factor for osteoporotic vertebral fractures. Physical growth in early life is related to bone health in later life, but the relationship of early growth versus vertebral size has been inconclusively studied. Utilizing the Northern Finland Birth Cohort 1966 with a 47-year follow-up, we investigated how physical growth in early life is associated with midlife vertebral dimensions. We obtained several physical growth parameters of 1) birth (gestational age, length, weight, BMI), 2) infancy and childhood (peak height velocity (PHV), peak weight velocity (PWV), adiposity peak (AP), adiposity rebound (AR)), and 3) puberty (BMI at growth spurt take-off (TO), PHV, height change). We also studied 4) the ages at which AP, AR, pubertal TO and pubertal PHV occurred. The outcome variable, vertebral cross-sectional area (CSA), was obtained from magnetic resonance imaging scans at the mean age of 46.7 years (n = 517). Sex-stratified linear regression analyses were used with adjustments for gestational age, smoking, and education. Birth length/weight/BMI, and adult height/weight/BMI were also used as covariates, depending on the model. According to our results, birth weight (p ≤ 0.006) and infant PWV (p ≤ 0.001) were positively associated with midlife vertebral CSA among both sexes. Length/height variables were associated with vertebral size only before including adult height in the models, and became non-significant thereafter. Among women, BMIs at birth, AP, AR, and pubertal TO were positively associated with midlife vertebral CSA (p < 0.05), whereas among men, only high BMI at AR was associated with large vertebral size (p = 0.028). Gestational age and timing of growth were not associated with future vertebral CSA. We conclude that early life weight gain is positively associated with midlife vertebral CSA, and suggest that adult height may mediate the effect of height gain on vertebral size.