On 2015 Mar 16, David Ayoub commented:

That Contreras et al1 failed to report an association between vitamin D deficiency (VDD) and fracture risk is surprising in light of nearly a century of research linking vitamin D to bone health. We would like to suggest that study design limitations might have obscured such a potential relationship.

While fractures are a well-known complication of rickets/osteomalacia their association with subclinical forms of deficiency is less well established. The prevalence of fractures in rickets varies but typically found in a minority of individuals suffering from the disease. In a review of vitamin D and skeletal health in children, Moon et al recently summarized 17 modern studies of rickets. Fractures were observed in 13 of these studies among 1,177 children. The average fracture prevalence was 6% with a range between 0% and 20%. Only 2 studies involved a similar aged subset as reported by Contreras. Agarwal and Gulati reported a 6% fracture prevalence in 10-13 y/o children in India. Narchi et al reported no fractures among 10-15 year olds with rickets in Saudi Arabia. A comparison of fracture rates among vitamin D sufficient vs. deficient individuals would therefore require an adequately large sample size ideally in prospective studies. Contreras’ alternative approach that compared VDD rates among those with fractures and non-injured cohorts introduces several potential confounders.

Bone failure, especially with minor trauma, is considered to be a stochastic event – one associated with probabilities – caused by the interplay of various extrinsic (force/load) and intrinsic (bone strength) factors. Intrinsic factors put genetically susceptible individuals at risk for fractures when exposed to the random forces of trauma. The genetic component has been reported to be the most important intrinsic factor associated with bone failure. Other risk factors include, but are not limited to decreased bone loading, illness, various endocrine and nutritional conditions, medications, physiologic alterations associated with bone turnover, and growth spurts. Designing a population-based study to isolate the role of VDD in fractures would be nearly impossible unless other major determinants that effect bone quality and fracture risk are controlled.

Contreras provides little detail of the nature of forces applied to the 100 fracture cases. Even if vitamin D deficiency was the major contributor to diminished bone strength, a force is still required to produce a fracture. Contreras’ fracture group could have had other confounding conditions that affected bone strength, including familial predispositions that adversely influenced bone quality. Since forces vary considerably it is certain that some fractures would occur regardless of bone strength. The fact that 63% of cases with fractures following “minor” trauma had inadequate vitamin D levels would raise concern that diminished bone strength may have contributed to injury in some children.

It is possible that Contreras’ control population, similarly suffering from VDD ubiquitous to the general population, also had compromised bone strength and differed only from controls by having the good fortune of avoiding a significant accident. It would have been more ideal to choose a control group strictly comprised of children suffering traumatic injuries but without fractures instead of a group that included various acute medical conditions. In this scenario, all subjects of both groups would have been exposed to physical forces that would have challenged bone integrity. It is possible that their control group of emergency room patients without fractures suffered from conditions associated with higher rates of VDD and thus neutralizing any potential difference in vitamin D status from the fracture group.

In addition to Ryan, there are several more studies that did report an association between vitamin D and fractures. Ruohola et al reported a strong association between lower vitamin D levels and stress fractures among 800 randomly selected and prospectively followed young military recruits. Leboff et al reported lower mean vitamin D levels in women with postmenopausal hip fractures compared to two groups of postmenopausal women prior to elective hip replacements. In a nested case-control study of 1200 female Naval recruits Burgi et al reported increased risk of stress fractures among those who were vitamin D deficient. There was approximately half the risk of stress fractures in women in the top vs. the bottom quintile of vitamin D concentration. While we agree that one cannot say that vitamin D status alone could predict who will fracture, these alternatively designed studies suggest that it can identify an at-risk population.

Lastly, but perhaps of greater importance is the extraordinary rate (63%) of inadequate vitamin D status among Contreras’ emergency room-attended population. This observation alone deserves the full attention of local physicians and public health officials and goes far beyond the original mission of their study.

REFERENCES 1. Contreras JJ, Hiestand B, O'Neill JC, Schwartz R, Nadkarni M. Vitamin D deficiency in children with fractures. Pediatr Emerg Care. 2014;30(11):777-781. 2. Moon RJ, Harvey NC, Davies JH, Cooper C. Vitamin D and skeletal health in infancy and childhood. Osteoporos Int. 2014;25(12):2673-2684. 3. Agarwal A, Gulati D. Early adolescent nutritional rickets. J Orthop Surg (Hong Kong). 2009;17(3):340-345. 4. Narchi H, El Jamil M, Kulaylat N. Symptomatic rickets in adolescence. Arch Dis Child. 2001;84(6):501-503. 5. Ferrari S, Chevalley T, Bonjour JP, Rizzoli R. Genetic determinants of bone microstructure and fracture risk in childhood. Bone. 2007;40(3)(suppl 1):S12. 6. Melamed ML, Kumar J. Low levels of 25-hydroxyvitamin D in the pediatric populations: prevalence and clinical outcomes. Ped Health. 2010;4(1):89-97. 7. Ryan LM. Forearm fractures in children and bone health. Curr Opin Endocrinol Diabetes Obes. 2010;17(6):530-534. 8. Ruohola JP, Laaksi I, Ylikomi T, et al. Association between serum 25(OH)D concentrations and bone stress fractures in Finnish young men. J Bone Miner Res. 2006;21(9):1483-1488. 9. LeBoff MS, Kohlmeier L, Hurwitz S, Franklin J, Wright J, Glowacki J. Occult vitamin D deficiency in postmenopausal US women with acute hip fracture. JAMA. 1999;281(16):1505-1511. 10. Burgi AA, Gorham ED, Garland CF, et al. High serum 25-hydroxyvitamin D is associated with a low incidence of stress fractures. J Bone Miner Res. 2011;26(10):2371-2377.

On 2015 Mar 16, David Ayoub commented:

That Contreras et al1 failed to report an association between vitamin D deficiency (VDD) and fracture risk is surprising in light of nearly a century of research linking vitamin D to bone health. We would like to suggest that study design limitations might have obscured such a potential relationship.

While fractures are a well-known complication of rickets/osteomalacia their association with subclinical forms of deficiency is less well established. The prevalence of fractures in rickets varies but typically found in a minority of individuals suffering from the disease. In a review of vitamin D and skeletal health in children, Moon et al recently summarized 17 modern studies of rickets. Fractures were observed in 13 of these studies among 1,177 children. The average fracture prevalence was 6% with a range between 0% and 20%. Only 2 studies involved a similar aged subset as reported by Contreras. Agarwal and Gulati reported a 6% fracture prevalence in 10-13 y/o children in India. Narchi et al reported no fractures among 10-15 year olds with rickets in Saudi Arabia. A comparison of fracture rates among vitamin D sufficient vs. deficient individuals would therefore require an adequately large sample size ideally in prospective studies. Contreras’ alternative approach that compared VDD rates among those with fractures and non-injured cohorts introduces several potential confounders.

Bone failure, especially with minor trauma, is considered to be a stochastic event – one associated with probabilities – caused by the interplay of various extrinsic (force/load) and intrinsic (bone strength) factors. Intrinsic factors put genetically susceptible individuals at risk for fractures when exposed to the random forces of trauma. The genetic component has been reported to be the most important intrinsic factor associated with bone failure. Other risk factors include, but are not limited to decreased bone loading, illness, various endocrine and nutritional conditions, medications, physiologic alterations associated with bone turnover, and growth spurts. Designing a population-based study to isolate the role of VDD in fractures would be nearly impossible unless other major determinants that effect bone quality and fracture risk are controlled.

Contreras provides little detail of the nature of forces applied to the 100 fracture cases. Even if vitamin D deficiency was the major contributor to diminished bone strength, a force is still required to produce a fracture. Contreras’ fracture group could have had other confounding conditions that affected bone strength, including familial predispositions that adversely influenced bone quality. Since forces vary considerably it is certain that some fractures would occur regardless of bone strength. The fact that 63% of cases with fractures following “minor” trauma had inadequate vitamin D levels would raise concern that diminished bone strength may have contributed to injury in some children.

It is possible that Contreras’ control population, similarly suffering from VDD ubiquitous to the general population, also had compromised bone strength and differed only from controls by having the good fortune of avoiding a significant accident. It would have been more ideal to choose a control group strictly comprised of children suffering traumatic injuries but without fractures instead of a group that included various acute medical conditions. In this scenario, all subjects of both groups would have been exposed to physical forces that would have challenged bone integrity. It is possible that their control group of emergency room patients without fractures suffered from conditions associated with higher rates of VDD and thus neutralizing any potential difference in vitamin D status from the fracture group.

In addition to Ryan, there are several more studies that did report an association between vitamin D and fractures. Ruohola et al reported a strong association between lower vitamin D levels and stress fractures among 800 randomly selected and prospectively followed young military recruits. Leboff et al reported lower mean vitamin D levels in women with postmenopausal hip fractures compared to two groups of postmenopausal women prior to elective hip replacements. In a nested case-control study of 1200 female Naval recruits Burgi et al reported increased risk of stress fractures among those who were vitamin D deficient. There was approximately half the risk of stress fractures in women in the top vs. the bottom quintile of vitamin D concentration. While we agree that one cannot say that vitamin D status alone could predict who will fracture, these alternatively designed studies suggest that it can identify an at-risk population.

Lastly, but perhaps of greater importance is the extraordinary rate (63%) of inadequate vitamin D status among Contreras’ emergency room-attended population. This observation alone deserves the full attention of local physicians and public health officials and goes far beyond the original mission of their study.

REFERENCES 1. Contreras JJ, Hiestand B, O'Neill JC, Schwartz R, Nadkarni M. Vitamin D deficiency in children with fractures. Pediatr Emerg Care. 2014;30(11):777-781. 2. Moon RJ, Harvey NC, Davies JH, Cooper C. Vitamin D and skeletal health in infancy and childhood. Osteoporos Int. 2014;25(12):2673-2684. 3. Agarwal A, Gulati D. Early adolescent nutritional rickets. J Orthop Surg (Hong Kong). 2009;17(3):340-345. 4. Narchi H, El Jamil M, Kulaylat N. Symptomatic rickets in adolescence. Arch Dis Child. 2001;84(6):501-503. 5. Ferrari S, Chevalley T, Bonjour JP, Rizzoli R. Genetic determinants of bone microstructure and fracture risk in childhood. Bone. 2007;40(3)(suppl 1):S12. 6. Melamed ML, Kumar J. Low levels of 25-hydroxyvitamin D in the pediatric populations: prevalence and clinical outcomes. Ped Health. 2010;4(1):89-97. 7. Ryan LM. Forearm fractures in children and bone health. Curr Opin Endocrinol Diabetes Obes. 2010;17(6):530-534. 8. Ruohola JP, Laaksi I, Ylikomi T, et al. Association between serum 25(OH)D concentrations and bone stress fractures in Finnish young men. J Bone Miner Res. 2006;21(9):1483-1488. 9. LeBoff MS, Kohlmeier L, Hurwitz S, Franklin J, Wright J, Glowacki J. Occult vitamin D deficiency in postmenopausal US women with acute hip fracture. JAMA. 1999;281(16):1505-1511. 10. Burgi AA, Gorham ED, Garland CF, et al. High serum 25-hydroxyvitamin D is associated with a low incidence of stress fractures. J Bone Miner Res. 2011;26(10):2371-2377.