Seasonal Variation of Bone Turnover?

To the Editor: We read with interest the report by Blumsohn et al.1 published in the July issue of the Journal. Based on cosinor analyses of a smaller longitudinal and a larger cross-sectional sample, the authors confirmed significant seasonal variations in serum vitamin D levels but were unable to detect any effect of season on serum markers of bone resorption. Because the authors mainly refer to two of our previous studies on circannual changes in bone turnover,2, 3 we feel obliged to comment on some aspects of their work1 and on the phenomenon of seasonal variability of bone turnover in general. When it comes to the assessment of temporal fluctuations in biological functions, most people would agree that data obtained from longitudinal studies prevail in significance over those acquired from cross-sectional surveys, no matter how big the latter may be. Including the most recent report in the Journal, a total of about 10 well-designed longitudinal studies investigating the effects of season on bone turnover have been published since 1986.1, 3-11 All of these studies show clear seasonality in serum 25(OH)vitamin D [25(OH)D] levels, with a nadir consistently occurring in winter (i.e., January-February). This fall in serum 25(OH)D levels coincides with a significant increase in bone turnover in some3, 5, 8, 10 but not all1, 4, 6, 7, 9, 11 studies. A careful analysis of the existing literature provides clues that may help to explain some of these apparent discrepancies. First, several of the negative studies used bone markers with recognized low sensitivity (e.g., serum total alkaline phosphatase, urinary hydroxyproline).4, 7, 11 As circannual changes in bone turnover are of comparatively small magnitude, these studies may have underestimated the effects of season on bone turnover simply by lack of methodological sensitivity. Second, population characteristics and composition, as well as sample size, seem worth a closer look. For example, in some studies, significant seasonality in bone turnover was observed in women but not in men.3, 10 There is some evidence that gender is an independent determinant of serum 25(OH)D levels.12 Also, compared with men, women in Northern Europe seem to require higher serum 25(OH)D levels to maintain normal serum parathyroid hormone (PTH) concentrations.13 These gender differences may explain, at least in part, why women are at greater risk for subclinical vitamin D deficiency and why studies including predominantly males1, 4, 7 were unable to show significant seasonal variations in bone turnover. Along these lines, it is interesting to note that all longitudinal studies reporting circannual changes in bone turnover markers also observed a significant seasonal variation in serum PTH levels.3, 5, 8, 10 Conversely, seasonal changes in serum PTH levels were always absent in the negative studies. Taken together, these observations support the hypothesis that transient secondary hyperparathyroidism on the grounds of subclinical vitamin D deficiency may be the driving force behind the winter increase in bone turnover. When directly comparing the longitudinal studies by Blumsohn et al.1 and Woitge et al.,3 several important differences should be considered. First, the report by Blumsohn et al.1 is comprised of only 12 individuals (6 males, 6 females) between the ages of 22 and 44 years. The study by Woitge et al.3 included 41 subjects (15 males, 26 females) between 25 and 80 years of age. The latter report3 clearly states that a significant circannual rhythm in bone turnover was observed in only 26 of their 41 individuals, and this phenomenon was absent in men. This negative observation alone would have accounted for 50% of the sample studied longitudinally by Blumsohn et al.1 Thus, if the longitudinal study by Woitge et al.3 had included a sample size of only six male and six female subjects, the chances of negative findings would have been greater than 70%. Second, as shown by the authors' and many other groups,14, 15 most markers of bone turnover are notorious for their high degree of (pre-analytical) variability. This kind of noise becomes a problem, particularly in smaller studies. Third, in addition to studying a larger sample and a wider age range, Woitge et al.3 also made an effort to generate a comparatively homogeneous population by assessing their subjects' lifestyle (e.g., smoking and alcohol use, calcium intake, physical activity) and anthropometric variables. The report by Blumsohn et al.1 makes no mention of such information being collected, which is particularly unfortunate in light of their small sample size and previous observations that age, gender, and menopausal status have a strong effect on the magnitude of seasonal changes in bone turnover. Thus, some of the differences in results from these and other reports may be attributable to variances in lifestyle and other confounding factors that were not taken into account in all of the studies. Finally, it is interesting to note that both the cross-sectional and the longitudinal studies by Blumsohn et al.1 reveal significant seasonal variations in serum vitamin D levels. These changes were similar or even identical to those observed in our and the studies of others. We have previously shown a potential relationship between the seasonal changes in vitamin D status and bone turnover,3 and have, more recently, confirmed these findings in a 2-year prospective, controlled intervention study (C Meier, HW Woitge, K Witte, R Ziegler, B Lemmer, and MJ Seibel, unpublished data, 2003). It makes little biological sense to us when Blumsohn et al. confirm circannual changes in an essential calciotropic hormone without seeing the effects of this change on the direct target, bone metabolism. In summary, seasonality in bone turnover is a subtle phenomenon that seems to be modulated by a number of cofactors such as, for example, age, gender, degree of and individual sensitivity to 25(OH)D deficiency, fluctuations in serum PTH levels, and in women, estrogen status. Apart from its impact on the clinical measurement of bone markers (which may or may not be significant), careful analyses of this phenomenon in sufficiently large and well-characterized populations may help to further elucidate the intricate play of bone metabolism and its regulation. We would certainly welcome such studies.