Intake of Vitamins B6 and C and the Risk of Kidney Stones in Women

Study Population

In 1976, 121, 701 female registered nurses between the ages of 30 and 55 yr residing in one of 11 U.S. states completed and returned the initial questionnaire and constitute the Nurses' Health Study (NHS) cohort (11). The cohort is followed by means of biennial mailed questionnaires that inquire about lifestyle practices and other exposures of interest, as well as the incidence of newly diagnosed disease. The population for the current analysis was limited to the women who answered the 1992 or the 1994 questionnaires, which inquired about a history of kidney stone disease, and who had completed at least one dietary questionnaire since dietary information was first collected in 1980.

We considered only cases that occurred during the 14-yr period between the return of the 1980 questionnaire and May 31, 1994. After we excluded women for whom the date of diagnosis could not be confirmed or fell outside the study period (n = 3233), 85,557 women with no history of kidney stones before the 1980 questionnaire remained in the study group.

Assessment of Diet

Participants were asked in 1980, 1984, 1986, and 1990 to complete semiquantitative food-frequency questionnaires on which they reported the average use of specified foods and beverages during the past year. The 1980 dietary questionnaire contained a list of 61 items, and the subsequent questionnaires contained approximately 130 items. Nutrient intake was computed from the reported frequency of consumption of each specified unit of food or beverage and from published data on the nutrient content of the specified portions (12,13). Beginning in 1980, information also was collected on supplemental vitamins, including vitamin C, either alone or as part of multivitamin preparations; a separate question on supplemental vitamin B6 intake was added starting in 1984. The multiple choice categories for supplemental vitamin C were (in mg/d): 0, 1 to 399, 400 to 700, 750 to 1250, and 1300 or more. Categories for supplemental vitamin B6 were (in mg/d): 0, 1 to 10, 10 to 39, 40 to 79, and 80 or more. The amounts of vitamins B6 and C in multivitamin preparations were determined by the brand, type, and frequency of reported use. The reproducibility and validity of the questionnaires in this cohort have been documented previously (12,13). The Pearson correlation coefficients for energy-adjusted vitamins B6 and C intake between the dietary records and the questionnaire were 0.54 and 0.73, respectively (12). After adjustment for the week-to-week variation of intake of vitamins B6 and C, the correlation coefficients were 0.58 and 0.75, respectively. After the exclusion of supplemental vitamin use, the correlation coefficients for vitamin B6 (0.54) and C (0.66) were slightly attenuated.

Nutrient values were adjusted for total energy intake using a regression model with total caloric intake as the independent variable and absolute nutrient intake as the dependent variable (13,14). Because total energy intake for a given person tends to be fixed within a narrow range, variations in nutrient intake are attributable largely to changes in composition of the diet, not the total amount of food consumed. Energy-adjusted values reflect the nutrient composition of the diet independent of the total amount of food consumed. In addition, energy adjustment reduces variation introduced by underreporting or overreporting of intake on the food-frequency questionnaire, thus improving the accuracy of nutrient measurements (13,14).

Assessment of Nondietary Factors

Body mass index (kg/m²) was calculated based on information on height which was provided in 1976 and weight that was provided on each subsequent follow-up questionnaire.

Follow-Up and Ascertainment of Cases

On the 1992 biennial questionnaire, women were asked whether they had ever been diagnosed with a kidney stone. On the 1994 questionnaire, they were asked about a new diagnosis of a kidney stone since 1992. If a kidney stone was reported to have occurred after the return of the 1980 questionnaire (when dietary information was first collected), we mailed the subject a supplementary questionnaire to confirm the diagnosis and to ascertain the date of first occurrence, symptoms, other relevant medical conditions and the type of stone if known. The response rate to the supplementary questionnaire was 92 percent. The validity of the subjects' reports was evaluated by review of the medical records from a random sample of 90 of the women who reported a kidney stone. The records confirmed the diagnosis in all but one of the cases.

Statistical Analyses

The study design was prospective with the dietary information collected before the onset of the kidney stone symptoms. For each participant, person-months of follow-up were counted from the date of the return of the 1980 questionnaire to the date of a kidney stone, death, or May 31, 1994, whichever occurred first. Information on exposures of interest from the 1980 questionnaire was updated in 1984, 1986, and 1990. We allocated person-months of follow-up according to exposure status (e.g., the category of vitamin use) at the start of each follow-up period. If information on vitamin intake was missing at the start of a time period, the subject was excluded from that time period. Subjects who reported a kidney stone on the biennial questionnaire but did not respond to the supplementary questionnaire were considered noncases in the analysis.

The relative risk—the incidence among women in a particular category of exposure divided by the corresponding rate in the comparison category—was used as the measure of association (15). Age-adjusted relative risks were calculated after stratification according to 5-yr age categories (15). The Mantel extension test was used to evaluate linear trends across categories of intake (16). We used a proportional hazards model to adjust for multiple risk factors simultaneously (17). The variables from the 1980 questionnaire in the proportional hazards model were updated in 1984, 1986, and 1990. The variables considered in these models were: vitamin B6 (five categories), vitamin C (five categories), age (in 5-yr categories), body mass index (seven categories), supplemental calcium (four categories), alcohol (six categories), and dietary intake of calcium, animal protein, potassium, sodium, sucrose, and total fluid (quintile groups). We selected these dietary variables based on our previous study, which demonstrated that these other dietary variables are related to risk of kidney stones (18). For all relative risks, we calculated 95% CI. All P values are two-tailed.

Vitamin B6

The characteristics of the cohort according to categories of energy-adjusted dietary vitamin B6 intake in 1986 are shown in Table 1. The 1986 values were selected to demonstrate quantitatively the boundaries of the categories and are representative of the ranges of intake during the different time periods for the cohort. However, for the analyses, the actual values as calculated from the food-frequency questionnaire were used for each respective time period. In 1986, 9.0% of the women reported consuming 10 mg or more each day of vitamin B6. Because the amount of dietary vitamin B6 intake did not differ substantially among the categories, it was the intake of supplemental vitamin B6 that resulted in assignment to higher categories. Mean daily intake of potassium, magnesium, supplemental calcium, and vitamin C increased with increasing intake of vitamin B6 (Table 1).

After adjusting for age only, there was no significant association between the risk of stone formation and vitamin B6 intake. However, after also adjusting for the intake of other dietary factors, women in the highest vitamin B6 category had a significantly lower risk of stone formation. The relative risk of stone formation in women in the highest (≥40 mg/d) category of vitamin B6 intake compared with the lowest category (<3 mg/d) was 0.66 (95% CI, 0.44 to 0.98) (Table 2). The results did not change when the raw vitamin B6 values were used instead of the energy-adjusted vitamin values. We also compared the risk for those women consuming ≥100 mg/d and found that the relative risk (RR 0.59; 95% CI, 0.32 to 1.08) was not significantly different from those taking 40 to 99 mg/d (RR 0.71; 95% CI, 0.43 to 1.17).

Vitamin C

In 1986, 7.9% of the cohort reported consuming 1000 mg or more each day of vitamin C. Similar to vitamin B6, the amount of dietary vitamin C did not differ substantially among the categories of intake; thus, the use of vitamin C supplements resulted in assignment to the higher categories. Mean daily intake of potassium, magnesium, supplemental calcium, and vitamin B6 increased with increasing intake of vitamin C (Table 3).

After adjusting for age, there was no significant association between the risk of stone formation and vitamin C intake (Table 4). The relative risk of stone formation for women in the highest category (≥1500 mg/d) of vitamin C intake compared with the lowest (<250 g/d) was 0.98 (95% CI, 0.65 to 1.47). There was no material change in these results after controlling for dietary factors or when the raw vitamin C values were used instead of the energy-adjusted vitamin valuves. In a multivariate model that included vitamin C as a continuous variable, for every 500 mg of vitamin C consumed daily, the relative risk of stone formation was 1.03 (95% CI, 0.95 to 1.11).

Vitamin B6

There are no randomized trials examining the impact of vitamin B6 supplementation on the risk of kidney stone formation. In uncontrolled studies, doses of 10 to 500 mg/d of pyridoxine were reported to decrease urinary oxalate (2,3,4) or stone recurrence rates (4,19) in calcium oxalate stone formers.

In contrast, investigators who prescribed 200 mg daily found no increase in urinary oxalate among stone formers but found a surprisingly significant 21% increase in urine oxalate among healthy subjects (20). In our observational study in men, we found no beneficial effect of vitamin B6 on the risk of incident stone formation, even for intakes 40 mg/d and higher (RR 0.91; 95% CI, 0.64 to 1.31) (5). The reason for the difference in results between men and women is unclear. We are unaware of any data that suggest oxalate metabolic pathways differ by gender.

Vitamin C

Short-term studies of the effect of large doses of oral vitamin C on urinary oxalate excretion have produced conflicting results (7,8,9). In one study of 39 healthy volunteers, urinary oxalate increased by 50% in subjects who received 1000 mg of vitamin C daily; no further increase was observed in those taking 9000 mg daily (9). Another study of 15 patients, who had recently undergone extracorporeal lithotripsy, found a 39% increase in urinary oxalate among those who received 1000 mg daily and a 100% increase with 2000 mg daily (8).

To determine whether the variability of results in previous studies was due to analytical errors, urinary oxalate levels were measured using modified ion chromatography after the oral administration of 1 to 10 g of vitamin C to healthy adults (7). Vitamin C added directly to urine resulted in a statistically significant but modest increase in measured oxalate, demonstrating that vitamin C was converted to oxalate during analysis. Thus, the increase in urine oxalate values that followed the ingestion of varying doses of vitamin C could be entirely accounted for by the conversion of vitamin C to oxalate during the analytical procedure. Therefore, the authors concluded that there was no true increase in urinary oxalate, despite the greatly increased vitamin C intake.

The results from studies that examined the association between vitamin C and risk of stone formation, and not only changes in urinary oxalate excretion, have not demonstrated an increase in risk. Case-control studies of the association between vitamin C intake and kidney stones found similar (21,22) or decreased (23,24) intake among the cases. A large cross-sectional study found that women currently consuming vitamin C supplements had a 10% lower risk of a self-reported history of kidney stones (10). However, since this was a cross-sectional rather than prospective study, this inverse association may be attributed to the discontinuation of vitamin C supplements by stone formers in response to medical advice. In our prospective observational study in men, we found no increased risk of incident stone formation, even for intakes of vitamin C ≥1500 mg/d (RR 0.78; 95% CI, 0.54 to 1.11) (5).

We avoided biased recall of diet by collecting vitamin intake data before the diagnosis of kidney stones. In addition, we adjusted for other dietary and nondietary risk factors for stone formation. Although we did not have information on the type of stone, we believe that more than 80% of the stones in our cohort were calcium oxalate (25). Because of the beneficial effects of large doses of pyridoxine on reducing oxalate excretion in patients with primary hyperoxaluria (26), it remains possible that a higher intake of vitamin B6 may have a positive effect in a subset of kidney stone patients with marked hyperoxaluria. However, we did not have sufficient information on urinary oxalate excretion to evaluate this issue.

The validity of our dietary questionnaire has been carefully documented, yet we recognize that in our study vitamin intake was not perfectly measured. Nevertheless, our findings of an inverse association with vitamin B6 cannot be explained by random misclassification. Moreover, most of the vitamin consumption in the high categories was from supplements that are well reported in this cohort. The range of intake of vitamin C was substantial, and the mean daily consumption in the highest category was more than 30 times greater than the U.S. recommended dietary allowance of 60 mg. Therefore, the lack of an association with vitamin C was not due to insufficient variation or magnitude of intake.

These results are generalizable to women, 34 yr of age or older, who have no history of kidney stones. In addition, given that the pathophysiology of stone formation is the same for both incident and recurrent kidney stones, it is likely that our results also apply to women who have a history of calcium oxalate stones. The results of this study of women and vitamin C intake are consistent with our previous findings in men. However, the inverse association between vitamin B6 intake and stone formation in women was not observed in men (5).

Large doses of vitamin B6 and C are taken relatively frequently by adults in the United States and may be beneficial for a variety of clinical conditions. Our results suggest that ≥40 mg/d of vitamin B6 intake may reduce the risk of stone formation in women. However, our findings for vitamin C, which have been consistent for women and men, do not support the practice of routine restriction of vitamin C to prevent kidney stones.