Vitamin D and Osteoarthritis
By Jane Kang, MD and Sharon L. Kolasinski, MD. Dr. Kang is a Rheumatology Fellow in the Division of Rheumatology at the University of Pennsylvania; Dr. Kolasinski is Associate Professor of Clinical Medicine, Interim Division Director and Fellowship Program Director in the Division of Rheumatology at the University of Pennsylvania. Drs. Kang and Kolasinski report no financial relationship to this field of study.
Both vitamin D deficiency and osteoarthritis (OA) are common problems in an aging population. Given the critical role of vitamin D in maintaining bone health and in calcium metabolism, a considerable number of laboratory and clinical studies have investigated the potential link between vitamin D and OA. There is currently no cure for OA and the identification of a preventive or disease-modifying treatment would be a great advance.
Background
Vitamin D deficiency (defined as a serum 25-hydroxyvitamin D level of less than 20 ng/ml) is common, particularly in elderly men and women where 40-100% in the U.S. and Europe are vitamin D deficient. Furthermore, more than half of postmenopausal women taking medication for osteoporosis are vitamin D insufficient (defined as a serum 25-hydroxyvitamin D level less than 30 ng/mg).1 While it has long been known that vitamin D deficiency causes rickets and osteomalacia, consequences of suboptimal levels of vitamin D also include an increased risk of osteopenia, osteoporosis and hip fractures.
Osteoarthritis is the most common form of arthritis and it is estimated that about 6% of adults ≥ 30 years of age have recurrent knee pain and radiographic evidence of osteoarthritis of the knee.2 OA also commonly affects the hips, small joints of the hands and the spine. The prevalence of osteoarthritis increases with age, obesity, knee injury, knee surgery and occupations that involve bending and lifting.3 Genetic factors also play a role. While damage to the articular cartilage is the hallmark of osteoarthritis, bony abnormalities such as osteophyte formation, subchondral sclerosis and cyst formation also occur as the result of altered osteoblast and osteoclast function. In addition, many develop weakness of the periarticular muscles and some become severely debilitated.
Mechanism of Action
Vitamin D is obtained mainly through de novo synthesis in the skin after ultraviolet B light exposure. Only about 20% of the daily requirement is acquired through dietary intake. Once vitamin D is obtained via synthesis or food, it is converted to 25-hydroxyvitamin D in the liver and circulates in the blood. This form of vitamin D needs to be converted to the active form, 1,25 dihydroxyvitamin D, in the kidney with the help of 25(OH)D3-1-a-hydroxylase. The active form is a lipid soluble hormone that signals the intestines to increase calcium and phosphorus absorption.4 Vitamin D deficiency not only affects calcium metabolism, it also influences osteoblasts, chondrocytes, matrix ossification, and bone density.5 Vitamin D binds to vitamin D receptors (VDR) on osteoblasts, the bone matrix producing cells, leading to maturation of osteoclasts (the cells that break down bone) and calcium mobilization from bone.6
While the exact biochemical pathways involving vitamin D and osteoblasts, osteoclasts and chondrocytes are not well understood, there is evidence that vitamin D may have a number of different ways of influencing these cells. For example, lack of signaling from VDRs promotes increased differentiation of osteoblast precursors into osteoblasts.6 In terminally differentiated osteoblasts, stimulation of VDRs may increase the activity of osteoblasts.6 In the absence of vitamin D, the balance between osteoblastic and osteoclastic function may be lost, leading to abnormal bone.
Because bone is clearly altered in osteoarthritis, vitamin D could affect the incidence or progression of osteoarthritis through its overall effects on bone, for instance, through changes in bone mineral density (BMD). In the past, it was suggested that OA and osteoporosis, characterized by low BMD, do not coexist. However, there have been recent studies that challenge that notion. Low BMD has been associated with increased risk of knee OA progression.7 Also, remodeled bone in OA may have less shock-absorbing capacity than normal bone.8 Therefore, the cartilage, already damaged by metalloproteinases (the most important group of destructive enzymes in the pathogenesis of OA), may be further mechanically damaged because it is supported by weakened bone. Hypothetically, then, vitamin D could play a role in the OA pathogenesis.
In addition to effects on osteoblasts and osteoclasts, vitamin D may influence the activities of chondrocytes. Chondrocytes produce the proteins that are building blocks for articular cartilage and the metalloproteinases that break down cartilage. Among other actions, 1,25-dihydroxyvitamin D may influence chondrocyte growth, differentiation or function in the growth plate.6 There is evidence that vitamin D increases proteoglycan production by mature chondrocytes,9 and that low 1,25-vitamin D levels increase metalloproteinase activity in vitro.10.11 The balance of cartilage formation and degradation is lost in OA, where cartilage loss is one of the hallmarks of the disease. Thus, the potential effects of vitamin D on cartilage provides another possible mechanism through which vitamin D might have an influence on OA. These sorts of observations have led investigators to explore the relationship between OA and vitamin D through a number of clinical studies.
Clinical Studies
Establishing an epidemiologic link between low serum vitamin D and progression of OA was one goal of a study by McAlindon and colleagues.12 These investigators sought to assess whether dietary intake of vitamin D and serum vitamin D levels would predict the incidence and progression of knee OA in a prospective observational study. Subjects were from the Framingham Heart Study and had anteroposterior (AP) weight-bearing knee radiographs performed between 1983-85 and 1992-93. Vitamin D intake was evaluated using a validated dietary questionnaire. Questionnaires that were incomplete (>12 items left blank) and those subjects who reported unlikely dietary intakes (<600 or >4000 kcal/d) were excluded. The presence of osteoarthritis was defined radiographically by a modified Kellgren and Lawrence (KL) scale as those having KL grade 2 or greater (defined as either definite presence of osteophytes with possible joint space narrowing or definite mild joint space narrowing with or without osteophytes). Baseline and follow up radiograph pairs were read in blinded fashion by both a radiologist and a rheumatologist. Controls were defined as those who did not have incident OA. Confounding factors such as supplement use, health status and BMD were also taken into account.
At baseline, 556 subjects were included in the study (mean age 70.3), of which 394 had no knee OA and 162 had existing disease, with 80 of those having bilateral knee OA. Over the course of the study, 75 knees developed OA and 62 knees showed progressive OA. Serum vitamin D levels were positively correlated with vitamin D intake (r = 0.24, p < 0.001). Incident knee OA that occurred after baseline was not associated with vitamin D intake or serum levels. However, for the middle and lowest tertiles for both vitamin D intake (middle: 170 - 347 IU/d, lowest: 3 - 170 IU/d) and serum vitamin D levels (middle: 27.0 - 33.0 ng/ml, lowest: 4.9 - 24.0 ng/ml), risk for progression of knee OA was increased about threefold. Low serum vitamin D was also associated with increased joint space narrowing (OR 2.3, CI 0.9 - 5.5) and osteophyte growth (OR 3.1, CI 1.3 - 7.5).
Individuals that did not have the appropriate radiographic studies were excluded, and this group was older (mean age 72.4, p < 0.001), had more men (43% vs 37%, p < 0.05), and was less active (p < 0.05). However, their body mass indices (BMIs), BMDs, serum vitamin D levels, vitamin D intake, and total energy intake were similar. Given that these subjects were older (and probably had the same or higher prevalence of OA as the subjects who were included in the analysis), had similar serum vitamin D levels and similar vitamin D intake, the results were not thought to have been biased toward the null.
A second epidemiologic study by Lane and colleagues hypothesized that vitamin D deficiency could be associated with the incidence of radiographic hip OA.13 Their study population included elderly white women (age > 65, mean age 71) who were participants in the Study of Osteoporotic Fractures (SOF). The SOF study excluded black women since they have a lower incidence of hip fractures, as well as women who were nonambulatory or had received bilateral hip replacements. Subjects were recruited from Baltimore, Minneapolis, Portland, and the Monongahela Valley near Pittsburgh. The study group was composed of 237 women who were randomly selected and did not differ significantly with the overall cohort in age, weight, bone density or vitamin D supplementation. They were included in the subgroup analysis of baseline serum vitamin D levels and hip radiographs performed at baseline, and follow ups were available for review. Subjects that did not have follow up radiographs (and were, therefore, excluded) were older, less healthy, less educated, more sedentary, had lower supplemental vitamin D intake and lower bone mineral densities. Hip radiographs with findings consistent with inflammatory arthritis, Paget's disease, total joint replacement or surgery for hip fracture were excluded.
The demographic characteristics, vitamin D supplement intake, overall health, number of sedentary hours per day, calorie expenditure per week and average serum 25- and 1,25-vitamin D levels were not significantly different between those who had incident hip OA and those that did not have OA at baseline or later in the study. Still, those with incident hip OA had higher bone mineral densities (p < 0.05). Individuals in the lowest (8 - 22 ng/ml) and middle (23 - 29 ng/ml) tertiles of serum 25-vitamin D levels had a threefold increase in incident joint space narrowing (lowest tertile: OR 3.34, CI 1.13 - 9.86, p = 0.029; middle tertile: OR 3.21, CI 1.06 - 9.68, p = 0.038) compared to the highest tertile group (30 - 72 ng/ml). However, serum 25-vitamin D levels were not associated with incident osteophytes. Interestingly, there was no significant association between 1,25-vitamin D and incident hip OA. Despite making an epidemiological link between vitamin D and OA, these results differed from those of McAlindon et al, which showed no association between incident OA and serum vitamin D levels but rather an association between vitamin D levels and the prevalence of OA.12 The fact that both studies found a link was intriguing but the study by Lane et al. only included elderly white women, making it difficult to generalize to the population as a whole. In addition, only one measurement of serum vitamin D was made in the Lane study and there was no information obtained about lifetime vitamin D intake. Since OA develops over decades, a single vitamin D level might not adequately reflect the role of vitamin D in its pathogenesis. Nonetheless, these results imply that vitamin D has a role in cartilage metabolism given its association with joint space narrowing, perhaps by affecting proteoglycan and metalloproteinase formation as discussed previously.
A subsequent study also explored the association of vitamin D levels with OA in a group of subjects with osteoporosis.14 Glowacki and colleagues studied postmenopausal women with osteoarthritis who were scheduled for total hip arthroplasty. Individuals were excluded if they had confounding medical conditions (renal insufficiency, malabsorption, gastrectomy, active liver disease, acute myocardial infarction, alcoholism and anorexia nervosa) or were taking medications that would interfere with bone mass. Sixty-eight Caucasian women were enrolled in the study with ages ranging from 46 to 89 years. Using a questionnaire, lifestyle, reproductive factors, dietary calcium intake and physical activity were assessed. BMD of the spine, proximal femur and total body were obtained with dual-energy x-ray absorptiometry. The BMDs for vertebrae with "moderately severe" osteoarthritis, disc-space narrowing or fracture were not included to avoid including artefactually high BMDs caused by degenerative disease of the spine. Serum 25-hydroxyvitamin D, 1,25-dihydroxyvitamin D, intact parathyroid hormone, osteocalcin, bone-specific alkaline phosphatase and urinary N-telopeptide levels were measured.
Glowacki et al. found that 25% (17 of the 68 included subjects) of postmenopausal women undergoing hip replacement for severe OA were osteoporotic, substantiating the fact that OA and osteoporosis are not mutually exclusive and, therefore, that osteoporotic fractures may not necessarily be decreased in those with OA. There were significant differences in age, years since menopause, and body weight between those with and without osteoporosis, but subjects were not significantly different in terms of body composition, calcium and alcohol intake, physical activity or estrogen use. Osteoporotic and non-osteoporotic individuals were significantly different in terms of smoking history, but those without osteoporosis had a higher number of pack-years. Those with low BMD also had significant increases in bone turnover markers: urinary N-telopeptide levels were 2.3 times higher, serum osteocalcin was 2.2 times higher and serum bone-specific alkaline phosphatase was 9% higher in women with osteoporosis compared to those who were not osteoporotic. These differences in bone turnover markers remained significant after adjustment for differences in age, age at menopause, and number of years since menopause. Additionally, 22% of all participants were vitamin D deficient (25-hydroxyvitamin D <37.5 nmol/L), with 4% having an increased level of serum parathyroid hormone (>6.8 pmol/L). Only two of the vitamin D deficient women had osteoporosis and, therefore, vitamin D levels were not associated with BMD (p = 0.32). The proportion of subjects with vitamin D deficiency and hyperparathyroidism was similar in both osteoporotic and non-osteoporotic groups. Osteoporosis was found during the entire postmenopausal period, contradicting the notion that women in the early postmenopausal stages do not get osteoporosis. In this study, vitamin D deficiency was not more prevalent in osteoarthritic women with low BMD than those with normal BMD, emphasizing the importance of assessing the presence of osteoporosis and vitamin D deficiency even in patients with osteoarthritis. While Glowacki's study showed that OA and osteoporosis can commonly coexist, Zhang and colleagues took this observation a step further in noting that, in those with low BMD and OA, not only is there a concern for fracture risk, but also of possible OA progression.7 A study done by Hart et al. found that low BMD at the hip was weakly related to OA progression,7 and women with fractures in the past seemed to be less likely to develop OA, independent of BMD.15 Unfortunately, the investigations done by Hart and Zhang et al. did not include evaluations of vitamin D status.
In a population-based survey done by Bischoff-Ferrari et al, this possible link between serum 25-hydroxyvitamin D levels and bone mineral density in knee OA was explicitly investigated.16 A total of 228 subjects with primary knee OA from the original and offspring cohorts of the Framingham study were classified as vitamin D deficient (25-OH vitamin D � 15 ng/ml), having hypovitaminosis D (25-OH vitamin D 16 - 32 ng/ml) or vitamin D replete (25-OH vitamin D > 32 ng/ml). Average BMD of these groups was assessed and adjusted for sex, age, BMI, knee pain, physical activity and disease severity. Individuals were considered to have knee OA if they had at least one knee with a KL grade 2 or more. Blinded review of the radiographs was complete by a bone radiologist and a rheumatologist. Included and excluded subjects did not differ in sex, BMI, physical activity, and disease severity but excluded subjects were younger (mean age 66.7) than those who were included (mean age 74.4). Only 35% of the subjects were men. Compared to men, women on average were 6 years older, had lower femoral neck BMD, were less physically active, and had lower serum 25-hydroxyvitamin D scores. In the entire study group, 15% were vitamin D deficient, 51% had hypovitaminosis D and only 34% were considered vitamin D replete. Furthermore, more women than men had vitamin D deficiency (34% men, 66% women) and more women than men had hypovitaminosis D (31% men, 69% women). Those who were vitamin D replete had lower BMIs and were older. The prevalence of knee pain in one knee was lower in those that were vitamin D replete compared to both the hypovitaminosis D and vitamin D deficient groups. The prevalence of bilateral knee pain was lower when compared to the hypovitaminosis D individuals. Higher BMDs were seen in the vitamin D replete (8.5% higher BMD, p = 0.02) and hypovitaminosis D (7.3% higher BMD, p = 0.02) groups compared to the vitamin D deficient participants, with the trend being statistically significant (p = 0.04). This trend was similar in both men and women, although the men had a greater increase in BMD compared to women. Bischoff-Ferrari and colleagues showed a significant positive correlation between 25-hydroxyvitamin D levels and BMD in those with knee OA, suggesting that supplementing these individuals with vitamin D can improve BMD.
One final epidemiologic study seemed to contradict the findings of earlier work and suggested that no significant link exists between vitamin D intake and OA.17 Felson and colleagues tried to confirm the association between 25-hydroxyvitamin D deficiency and increased joint space narrowing in OA using two longitudinal cohorts, the Framingham Osteoarthritis Study (FOS) and the Boston Osteoarthritis of the Knee Study (BOKS). The FOS group consisted of the offspring of the original Framingham cohort, whereas the study done by McAlindon et al. emphasized the original Framingham cohort.12 The BOKS group included subjects recruited from a Veterans Affairs Medical Center. For the FOS group, weight bearing anteroposterior (AP) and lateral knee radiographs were obtained. In the BOKS group, fluoroscopically positioned semiflexed posteroanterior (PA) and lateral radiographs and magnetic resonance imaging (MRI) of the more painful knee were done. For both groups, worsening OA was defined as joint space loss in the tibiofemoral joint as seen on radiographs. In the BOKS study, loss of cartilage was also assessed via MRI. Adjustments were made for age, BMI, sex, and baseline OA.
There were 715 and 277 subjects included from the FOS and BOKS groups, respectively. In the FOS group, the mean age was 53.1 years, 53% were women, and 87% did not have OA at the beginning of the study (K&L score <2). After an average of 9.5 years, 20.3% of those without OA at baseline in the FOS group developed joint space narrowing. The subjects in the BOKS were older (mean age 66.2), heavier (mean BMI 31.2), and had more OA at baseline (21% had K&L grade <2 at baseline) compared to the FOS group, with 23.6% of knees with worsening OA. Of the 715 subjects in the FOS study, the mean 25-hydroxyvitamin D level was 20 ng/ml at baseline. In the 277 subjects in the BOKS study, the average 25-hydroxyvitamin D level was also 20 ng/ml at baseline. The authors concluded that there was no association between baseline vitamin D deficiency and worsening OA by radiographs or MRI. Interestingly, while those with vitamin D deficiency had a slightly decreased risk of worsening OA compared to those with higher serum 25-hydroxyvitamin D, it was not statistically significant.
Given the complexity of the potential relationship between vitamin D intake and osteoarthritis, it is perhaps not surprising that epidemiological studies have failed to establish a clear link. One additional avenue of investigation has focused on possible shared genetics. It has been suggested that mutations or polymorphisms of the COL2A1 or VDR gene, both on chromosome 12q, may be associated with OA. COL2A1 encodes the gene for type II collagen, which has been linked to some rare disorders of cartilage. Baldwin and colleagues investigated a possible genetic link between the VDR/COL2A1 locus and hand and knee OA through the original and offspring cohorts of the Framingham Osteoarthritis Study.
In this study, hand or knee radiographs were evaluated by a single reviewer. For the hand radiographs, the KL score from 10 distal interphalangeal, 8 proximal interphalangeal, 10 metacarpophalangeal, and 2 carpometacarpal joints were added together and adjusted for age and sex. The percentage of hand joints with a KL score of 2 or more was calculated. For knee radiographs, the KL grading was also used and sums of osteophytes (over 4 sites) and joints space narrowing (over 2 sites) were adjusted for age, body mass index (BMI) and history of knee injury or operation. Microsatellite markers bordering the VDR-COL2A1 gene were tested.
The average age of those in the study was 60 with equal percentages of male and female; 296 extended families were included. The investigators found no association between the COL2A1/VDR locus and hand/knee OA and suggested that mutations of polymorphisms in the VDR and COL2A1 genes are not a significant cause for OA. While this study implied that the COL2A1 locus had no significant role in OA, the importance of environmental influences, other genetic factors or the effects of these genes in different populations was not evaluated.
Conclusion
Epidemiological studies have failed to consistently link vitamin D to the incidence or prevalence of osteoarthritis. No randomized, prospective trials have been performed and, therefore, no conclusions can be drawn about supplementation with vitamin D to prevent or slow osteoarthritis. However, the current evidence does not suggest that this is likely to be a fruitful intervention. Nonetheless, further investigation is warranted given the complex and essential role of vitamin D in maintaining bone health. It is clear, however, that osteoarthritis and osteoporosis can occur in the same individual.
Recommendation
Lower serum vitamin D levels are associated with lower BMDs in those with or without OA. Numerous studies have shown that vitamin D supplementation can improve bone mineral density. Therefore, it is reasonable to provide supplements to vitamin D deficient or insufficient patients to improve bone density. While some clinical studies show a possible association with vitamin D and OA incidence or disease progression, the results are inconsistent and vitamin D cannot, at this time, be considered as a preventive or disease modifying agent in osteoarthritis.
References
1. Holick MF. Vitamin D deficiency. N Engl J Med. 2007;357(3):266-281.
2. Hunter DJ, Felson DT. Osteoarthritis. BMJ. 2006;332(7542):639-642.
3. Felson DT. Clinical practice. Osteoarthritis of the knee. N Engl J Med. 2006;354(8):841-848.
4. Arnson Y, et al. Vitamin D and autoimmunity: new aetiological and therapeutic considerations. Ann Rheum Dis. 2007;66(9):1137-1142. Epub 2007 Jun 8.
5. McAlindon T, Felson DT. Nutrition: risk factors for osteoarthritis. Ann Rheum Dis. 1997;56(7):397-400.
6. St-Arnaud R. The direct role of vitamin D on bone homeostasis. Arch Biochem Biophys. 2008;473(2):225-230. Epub 2008 Apr 6.
7. Schwartz Z, et al. Direct effects of transforming growth factor-beta on chondrocytes are modulated by vitamin D metabolites in a cell maturation-specific manner. Endocrinology. 1993;132(4):1544-1552.
8. Dean DD, et al. Vitamin D regulation of metalloproteinase activity in matrix vesicles. Connect Tissue Res. 1996;35(1-4):331-336.
9. Schmitz JP, et al. Vitamin D3 regulation of stromelysin-1 (MMP-3) in chondrocyte cultures is mediated by protein kinase C. J Cell Physiol. 1996;168(3):570-579.
10. Zhang Y, et al. Bone mineral density and risk of incident and progressive radiographic knee osteoarthritis in women: the Framingham Study. J Rheumatol. 2000;27(4):1032-1037.
11. Radin EL, Rose RM. Role of subchondral bone in the initiation and progression of cartilage damage. Clin Orthop Relat Res. 1986;213:34-40.
12. McAlindon TE, et al. Relation of dietary intake and serum levels of vitamin D to progression of osteoarthritis of the knee among participants in the Framingham Study. Ann Intern Med. 1996;125(5):353-359.
13. Lane NE, et al. Serum vitamin D levels and incident changes of radiographic hip osteoarthritis: a longitudinal study. Study of Osteoporotic Fractures Research Group. Arthritis Rheum. 1999;42(5):854-860.
14. Glowacki J, et al. Osteoporosis and vitamin-D deficiency among postmenopausal women with osteoarthritis undergoing total hip arthroplasty. J Bone Joint Surg Am. 2003;85-A(12):2371-2377.
15. Hart DJ, et al. The relationship of bone density and fracture to incident and progressive radiographic osteoarthritis of the knee: the Chingford Study. Arthritis Rheum. 2002;46(1):92-99.
16. Bischoff-Ferrari HA, et al. Positive association between serum 25-hydroxyvitamin D level and bone density in osteoarthritis. Arthritis Rheum. 2005;53(6):821-826.
17. Felson DT, et al. Low levels of vitamin D and worsening of knee osteoarthritis: results of two longitudinal studies. Arthritis Rheum. 2007;56(1):129-136.
Jane Kang, MD, and Sharon Kolasinski, MD. Vitamin D and Osteoarthritis. Alter Med Alert. 20088:11;85-90.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content