Vitamin C Supplement Use and Bone Mineral Density in Postmenopausal Women

Authors

  • Deborah J. Morton,

    1. Department of Family and Preventive Medicine. University of California, San Diego, La Jolla, California, USA
    Search for more papers by this author
  • Elizabeth L. Barrett-Connor,

    Corresponding author
    1. Department of Family and Preventive Medicine. University of California, San Diego, La Jolla, California, USA
    • Department of Family and Preventive Medicine, Division of Epidemiology, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093–0607, USA
    Search for more papers by this author
  • Diane L. Schneider

    1. Department of Medicine, University of California, San Diego, La Jolla, California, USA
    Search for more papers by this author

Abstract

Vitamin C is known to stimulate procollagen, enhance collagen synthesis, and stimulate alkaline phosphatase activity, a marker for osteoblast formation. Studies of dietary vitamin C intake and the relation with bone mineral density (BMD) have been conflicting, probably because of the well-known limitations of dietary nutrient assessment questionnaires. The purpose of this study was to evaluate the independent relation of daily vitamin C supplement use with BMD in a population-based sample of postmenopausal women. Subjects were 994 women from a community-based cohort of whom 277 women were regular vitamin C supplement users. Vitamin C supplement use was validated. Daily vitamin C supplement intake ranged from 100 to 5000 mg; the mean daily dose was 745 mg. Average duration of use was 12.4 years; 85% had taken vitamin C supplements for more than 3 years. BMD levels were measured at the ultradistal and midshaft radii, hip, and lumbar spine. After adjusting for age, body mass index (BMI), and total calcium intake, vitamin C users had BMD levels approximately 3% higher at the midshaft radius, femoral neck, and total hip (p < 0.05). In a fully adjusted model, significant differences remained at the femoral neck (p < 0.02) and marginal significance was observed at the total hip (p < 0.06). Women taking both estrogen and vitamin C had significantly higher BMD levels at all sites. Among current estrogen users, those also taking vitamin C had higher BMD levels at all sites, with marginal significance achieved at the ultradistal radius (p < 0.07), femoral neck (p < 0.07), and total hip (p < 0.09). Women who took vitamin C plus calcium and estrogen had the highest BMD at the femoral neck (p = 0.001), total hip (p = 0.05), ultradistal radius (p = 0.02), and lumbar spine. Vitamin C supplement use appears to have a beneficial effect on levels of BMD, especially among postmenopausal women using concurrent estrogen therapy and calcium supplements.

INTRODUCTION

RECENT IN vitro studies of animal and human osteoblasts have shown ascorbic acid (vitamin C) to be a vital component in the biology of bone cell formation and resultant bone mass. Vitamin C is known to enhance collagen synthesis, a precursor of bone matrix mineralization, to stimulate the overall process of osteoblast differentiation including the expression of alkaline phosphatase, a marker for osteoblast formation, and to increase the rate of osteoclast formation and increase the life span of osteoclasts and their precursors.(1–4) These results suggest a potentially beneficial role for vitamin C in the prevention of low bone mineral density (BMD). These observations could be important because vitamin C is the most common single nutritional supplement used in the United States.(5)

The results of studies that examined the relationship of dietary and/or supplemental vitamin C intake with BMD have been inconsistent. Some examined only dietary sources of vitamin C and did not consider supplemental intake.(6–7) Others examined total intake but did not determine whether vitamin C supplement users differ from nonusers,(8–10) or whether the effect was additive or interactive with calcium supplementation or estrogen use.

The purpose of the present study was to examine the independent relation between vitamin C supplement use and BMD at five skeletal sites in a community-dwelling population of postmenopausal women who had a high prevalence of vitamin C, calcium, and estrogen use.

MATERIALS AND METHODS

Between 1972 and 1974, 82% of all adult residents of the southern California community of Rancho Bernardo (n = 6339) were surveyed to estimate the prevalence of heart disease risk factors.(11,12) From February 1988 to November 1991, 80% of surviving noninstitutionalized postmenopausal women from the Rancho Bernardo cohort participated in a study of osteoporosis.(13) All were white, ambulatory, and gave written informed consent.

Standardized questionnaires were used to obtain information about alcohol intake, tobacco use, physical activity, disease history, reproductive status, medication, and nutritional supplement use. Information on all nutritional supplements and medications, including current use, dose, and duration, were obtained and validated by a nurse who examined pills and prescriptions brought to the clinic for that purpose. Height and weight were measured with the participant wearing light clothing and no shoes. Body mass index (BMI) was calculated as weight (kg) divided by height (m2).

BMD, defined as total bone mineral content (g) divided by the area (cm2) was measured at the ultradistal radius and midshaft radius of the nondominant arm using single photon absorptiometry (Lunar model SP2B; Lunar, Madison, WI, USA). Ultradistal radius BMD was defined as the lowest mean BMD of four adjacent scanned lines from a total of 10 lines. Midshaft radius BMD was the mean of four scanned lines at the proximal one-third of the radius. The Lunar bone densitometer was calibrated daily, using a phantom or calibration standard. Measurements were maintained within the manufacturer's precision standards of ≤7.0% for the ultradistal radius and 1.0% for the midshaft radius. Femoral neck, total hip, and lumbar spine BMD were measured by dual-energy X-ray absorptiometry (DXA; Hologic QDR-1000; Hologic, Inc., Waltham, MA, USA). Total hip BMD included the femoral neck, greater trochanter, and inter-trochanteric region. Spine BMD included lumbar verte-brae 1–4. Axial scans were standardized daily against a phantom. Precision error of this instrument was approximately 1%.

All BMD levels were distributed normally making transformations unnecessary. The direct method was used to calculate age adjustment using all 994 women. The Mantel Haenszel statistic with a two-tailed test of significance (α ≤ 0.05) was calculated to determine statistically significant differences in adjusted proportions. The SAS generalized linear model procedure was used to compare BMD levels stratified by vitamin C use, dose, and estrogen and/or calcium supplement use before and after adjusting for covariates.(14) Pearson correlation coefficients were calculated to test the relation of vitamin C dose and duration with BMD levels. Significant differences in mean values were determined using 95% CIs and p values with a two-tailed test of significance.

RESULTS

Participants in the study were 994 postmenopausal women, aged 50–98 years, with an overall mean age of 72 years. These women were an average of 26 years postmenopausal. Twenty-eight percent (n = 277) were regular daily users of vitamin C supplements. Compared with nonusers, vitamin C users were significantly more likely to currently use estrogen and take calcium and multivitamin supplements (Table 1). They also were significantly more likely to have physician-diagnosed (self-reported) osteoarthritis and significantly less likely to be taking thiazides. They were similar to nonusers of vitamin C with respect to age, BMI, use of oral steroids, physical activity, cigarette smoking, and alcohol use.

Table Table 1.. Age-Adjusted Sample Characteristics by Vitamin C Supplement Use in Women from Rancho Bernardo, CA, from 1988 to 1991
original image

Daily vitamin C supplement users had significantly higher BMD levels (p < 0.02) at the midshaft radius, femoral neck, and total hip when compared with nonusers after adjustment for age, BMI, and calcium supplement use (Table 2). A similar pattern was observed at the lumbar spine (p < 0.09). Differences in BMD levels remained at the femoral neck (p < 0.02) and total hip (p < 0.06) when all covariates were added to the model. Multiply adjusted BMD levels were 4.1% higher at the femoral neck in women using vitamin C supplements. Interactions between vitamin C use and multivitamin use, calcium supplements, and current estrogen, as well as calcium supplements and multivitamin use were not significant in any of these models.

Table Table 2.. Mean BMD Levels (95% CIs/p Values) Stratified by Current Vitamin C Supplement Use in Women from Rancho Bernardo, CA, from 1988 to 1991
original image

Dose of vitamin C supplements ranged from 70 to 5000 mg/day with a mean of 745 mg and a median of 500 mg/day. Fifty-six percent of vitamin C users took 500 mg/day, and 22% took 1000 mg/day. Two women used a vitamin C dose between 500 and 1000 mg/day and were excluded from the dose analysis. BMD levels stratified into three groups, nonusers, users of less than or equal to 500 mg/ day (n = 188 or 68% of vitamin C users), and users of 1000+ mg/day (n = 85 or 30.7% of vitamin C users), showed a pattern of increased bone mass at all sites and significantly higher BMD (p < 0.05) at the ultradistal radius for women taking the highest doses. A significant linear trend across the three groups was observed at the same site (Fig. 1).

Figure FIG. 1.

Mean BMD levels (adjusted for all covariates: age, BMI, exercise, alcohol, current smoking, calcium supplements, multivitamin supplements, thiazides, thyroid hormone, oral steroids, and physician diagnosed osteoarthritis; 95% CIs) stratified by dose (note that only 2 women used vitamin C at a dose level between 500 and 1000 mg) of current vitamin C supplement use in women from Rancho Bernardo, CA from 1988 to 1991.

Mean duration of use of vitamin C supplements was 12.4 years. Seventy-eight percent of vitamin C users had been taking vitamin C supplements for 5 or more years, and 56% had been taking vitamin C for 10 or more years. BMD levels were not significantly correlated with duration of use nor were there any differences when users were stratified into two groups, less than 10 years (n = 119, 43.6%) versus 10 or more years (n = 154, 56.4%) of use (data not shown).

Current users of both estrogen and vitamin C supplements had significantly higher BMD levels at all sites compared with women not taking estrogen and not taking vitamin C (Fig. 2). The difference in BMD levels ranged from 8.4% at the femoral neck to 16.6% at the ultradistal radius. Among current estrogen users, those taking vitamin C had higher BMD levels at all sites with marginal significance at the ultradistal radius (p < 0.07, 7.7%) and femoral neck (p < 0.07, 4.3%). A similar pattern also was observed among nonestrogen users, though the differences were not statistically significant.

Figure FIG. 2.

Mean BMD levels (adjusted for all covariates: age, BMI, exercise, alcohol, current smoking, calcium supplements, multivitamin supplements, thiazides, thyroid hormone, oral steroids, and physician diagnosed osteoarthritis; 95% CIs) stratified by current estrogen use and current vitamin C supplement use in women from Rancho Bernardo, CA from 1988 to 1991.

Further stratification of vitamin C supplement use into four groups, no estrogen/no calcium supplements (n = 415), calcium supplements only (n = 206), estrogen only (n = 182), and estrogen plus calcium supplements (n = 191), showed consistently at all five sites that women taking vitamin C and calcium supplements plus estrogen had the highest bone mass (Fig. 3). These differences were significant at the femoral neck (p = 0.0001), total hip (p = 0.05), and ultradistal radius (p = 0.02) compared with women taking calcium supplements plus estrogen but not vitamin C.

Figure FIG. 3.

Mean BMD levels (adjusted for all covariates: age, BMI, exercise, alcohol, current smoking, multivitamin supplements, thiazides, thyroid hormone, oral steroids, and physician diagnosed osteoarthritis; 95% CIs) stratified by current estrogen, calcium, and vitamin C supplement use in women from Rancho Bernardo, CA from 1988 to 1991.

DISCUSSION

Among these community-dwelling postmenopausal women, 28% were regular daily users of vitamin C supplements, a prevalence slightly lower than the 34% used by postmenopausal women reported from U.S. national data,(15) and the 36% reported from the Seattle site of the Fracture Intervention Trial.(16) Sowers and colleagues(8) reported more than one-third of the women in the rural Iowa study used some form of nutritional supplements, which included vitamin C. In a study of bone mass in Mexican American postmenopausal women,(17) data from Scotland,(9,10) and the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial,(7) prevalence of vitamin C supplement use was not reported separately.

Women from Rancho Bernardo, CA, who were vitamin C supplement users had significantly higher BMD levels at the femoral neck (p < 0.02) while controlling for all covariates. Significantly higher bone mass at the midshaft radius (p < 0.02) and total hip (p < 0.02) was observed while controlling for age, BMI, and use of calcium supplements. Differences in BMD levels ranged from 2.5% to 4.1% higher than among the nonusers.

In general, these results are in accord with two previous studies that have examined the relation of BMD and dietary vitamin C(6,7) and found positive associations at the ultradistal and distal radii, and at the femoral neck and total hip, respectively. However, one of these studies was among adolescents,(6) a period of bone formation as opposed to bone loss, and the other among women who were less than 10 years postmenopausal.(7) Results from the present study are similar to results from one study that accounted for both dietary and supplemental vitamin C intake by Sowers et al.,(8) who found a nonsignificant positive association of total vitamin C intake at the midshaft radius. In contrast, Wang et al.(17) found dietary but not supplemental vitamin C or total vitamin C intake to be marginally positively (p < 0.07) associated with BMD levels; the most significant differences were at the femoral neck, as in the present study. However, these results are difficult to interpret because the number of women using vitamin C supplements was not reported nor was a stratified analysis of BMD across dose levels conducted.

We found no association of vitamin C supplement use with bone mass at the lumbar spine, in contrast to New et al.(9) who examined premenopausal women and found among those in the highest quartile of total intake of vitamin C, lumbar spine BMD levels were significantly higher; vitamin C supplement users were not analyzed separately.

The daily dose of vitamin C from supplements among women from Rancho Bernardo, CA, was substantially higher (mean = 745 mg; median = 500 mg) than either vitamin C from supplement use, dietary intake, or total vitamin C dose reported in other studies, as well as the national average of non-Hispanic white women over 50 years old (94–112 mg).(18) Hall and Greendale(7) reported dietary vitamin C intake as 140 mg/day among the PEPI women. Leveille et al.(16) found mean dose from supplements was 295 mg/day and total intake was 407 mg/day. New and colleagues(9) reported total mean intake as 126 mg/day, and Wang et al.(17) found a dose of 139 mg/day for dietary vitamin C and 294 mg/day for vitamin C supplement use. We did not find any significant correlations between vitamin C dose and BMD levels at any site or when stratified by a median split (<500 mg vs. 500+ mg), possibly because the majority of women were taking 500 mg or more (87%). Further stratification showed a significant association of vitamin C dose and bone density at the ultradistal radius among women whose daily supplement intake was 1000 mg or more when compared with nonusers. This finding is in accord with the observation that 75% or more of vitamin C is absorbed at an intake of 180 mg of vitamin C, and pharmacokinetic studies show the intestine has an absorption capacity of approximately 3 g/day of ascorbic acid before the kidney effectively eliminates the excess.(5) Therefore, women taking higher doses would be expected to have higher serum vitamin C levels presumably enhancing the potential bone-preserving effects.

We found no association of duration of vitamin C supplement use and BMD at any site, in contrast to Leveille and colleagues(16) who reported that women who took high doses of vitamin C supplements for at least 10 years had higher BMD levels at the femoral neck. The lack of any strong and consistent associations of dose/duration of vitamin C supplement use and bone mass in Rancho Bernardo, CA, may be caused by a lack of variation in the distribution of the dose/duration variables. Most vitamin C users were taking relatively high doses for many years.

Leveille et al.(16) reported that among women aged 55–64 years who never used estrogen, femoral neck BMD was higher among those who had taken vitamin C supplements for longer than 10 years and suggested that vitamin C might be most beneficial during the early postmenopausal years, and estrogen use might obscure the more moderate effects of vitamin C. We did not find that to be true in our data. Rancho Bernardo women currently taking estrogen and vitamin C supplements had the highest bone mass at all five sites (p < 0.05), suggesting an additive effect. Among current estrogen users, vitamin C supplement users had marginally higher BMD levels at the ultradistal radius and femoral neck. Among nonestrogen users, BMD levels were slightly higher for users of vitamin C supplements. It is difficult to compare our estrogen results with Wang et al.,(17) Hall and Greendale,(7) or Sowers et al.(8) because they did not analyze the relation of estrogen plus vitamin C use to bone mass; they reported only the effect of estrogen use as a confounding variable in their regression models. Additionally, subjects in the study from New and colleagues(9) were all premenopausal and not using estrogen.

Recently, Nieves et al.(19) reviewed 20 clinical trials of estrogen therapy alone or in combination with added dietary calcium and found that higher calcium intake potentiated the beneficial effects of estrogen on bone at the hip, spine, and forearm. Data from the present study add a new dimension to the Nieves et al.(19) results because Rancho Bernardo women taking vitamin C plus calcium and estrogen had 13.5% higher BMD at the wrist, 9.5% higher BMD at the femoral neck, and 6.0% higher BMD at the total hip.

One limitation of this study is that we did not obtain serum levels of vitamin C. Data from the second National Health and Nutrition Examination Survey (NHANES II) have shown serum vitamin C levels in supplement users to be significantly lower in smokers compared with nonsmokers.(15) Although there were relatively few current smokers in the Rancho Bernardo cohort, there were more current smokers among the vitamin C supplement users compared with nonusers (11.2% vs. 9.8%). The associations reported here were adjusted for cigarette smoking, but smokers taking vitamin C supplements may have had significantly lower serum vitamin C levels not accounted for by these analyses.

Misclassification among the nonusers of vitamin C may have occurred in the present study, because 40% of Rancho Bernardo women were taking multivitamin preparations, which generally contain some vitamin C. However, the majority of the multivitamins used contained low to moderate levels of vitamin C (35–90 mg). To reduce the effect of possible confounding, the variable “multivitamin use” was included in all multivariate models. If misclassification did occur, it would tend to weaken rather than overestimate the observed associations of vitamin C use and bone mass.

A major strength of the present study is that current vitamin C supplement use and dosage were validated by clinic nurses. Supplement use in all other studies referenced was determined by recall. Nevertheless, although prevalence of vitamin C supplement use by Rancho Bernardo women was only slightly lower than national data, dose levels were substantially higher than the national average of reported use. These data, concordant with several other studies, support the thesis that regular daily use of vitamin C supplements may help increase bone mass in postmenopausal women at the femoral neck and total hip. Vitamin C alone cannot replace the documented benefits of estrogen replacement therapy and calcium supplements, but appears to have an additive effect. The optimal dose cannot be determined from the present report, but the highest BMD levels were observed among women taking 1000 mg/day or more.

Acknowledgements

This work was supported by the National Institute on Aging grant AG 07181.

Ancillary