Elderly women in northern New England exhibit seasonal changes in bone mineral density and calciotropic hormones

A seasonal component to bone loss has been reported in some postmenopausal women. We hypothesized that elderly women in northern New England would be at high risk for bone loss during winter because of their age, diet and lack of sunlight. Eighteen elderly but healthy women (mean age 77 years) started a 2-year observational study in a remote area of northwestern Maine (Greenville: 45.5°N latitude). Fifteen women completed the study. Bone mineral density of the spine (L-BMD) and hip (F-BMD) and biochemical markers of bone turnover were measured biannually. In vitro photo-conversion of provitamin D 3 to previtamin D 3 was determined in the winter and spring. Initial calcium and vitamin D intake averaged 700 ± 72 mg/day and 6.2 ± 1.2 μg/day, respectively. During the first year L-BMD dropped 4.2% ( P = 0.002) while F-BMD dropped 2.4% ( P = 0.09), primarily because of steep declines during the fall and winter (August to February: L-BMD: 3.6%, P = 0.001; F-BMD: 3.0%, P = 0.04). In that 6-month period, serum 25 hydroxyvitamin D (25(OH)D) fell 13 ± 6% (P = 0.06) and PTH rose 27 ± 11% ( P = 0.01). Additionally, there was no detectable in vitro conversion of provitamin D to previtamin D over 8 h of one sunny winter day. In the summer, PTH and 25(OH)D reverted to basal levels and significant in vitro photoconversion of provitamin D to previtamin D was detected. In the 2nd year of the study, vitamin D consumption increased (+2.0 ± 1.2 μg/day, P = 0.03 vs. baseline), L-BMD increased slightly (+1.8%, P = 0.05) and F-BMD did not change (+0.5%, P = NS). Again, changes in BMD and vitamin D were seasonal: L-BMD and F-BMD were constant during fall and winter but both sites showed increases during summer (L-BMD: +1.7%, P = 0.04, F-BMD: +1.6%, P = 0.25). In the second winter, serum 25(OH)D fell nearly 20% and PTH rose 17%. Increased dietary consumption of vitamin D was positively correlated with changes in F-BMD at 18 months ( r = 0.61, P = 0.02) and resulted in slightly greater serum 25(OH)D concentrations during the second winter than the first. The difference in serum 25(OH)D between the first and second winter was the strongest predictor of lumbar bone accretion during the second year of the study ( r = 0.59, P = 0.04). In this 24-month observational study, significant seasonal changes in BMD, 25(OH)D and PTH were reported. Winter was the time of lowest serum 25(OH)D and the greatest decline in measurable bone mass, while summer was associated with modest increases in seru