Incidence of Hypercalciuria and Hypercalcemia During Vitamin D and Calcium Supplementation in Older Women

John Christopher Gallagher, MD Lynette M. Smith, MSc; Vinod Yalamanchili, MD

Disclosures

Menopause. 2014;21(11):1173-1180. 

In This Article

Results

One hundred sixty-three women were randomized, 147 women completed the protocol, 5 women were lost to follow-up, and 11 women discontinued the study medications. Those women who discontinued the supplements came in for a final visit and were included in the analysis (Figure 1, Supplemental Digital Content 1).

Figure 1.

Incidence of hypercalcemia compared with placebo and vitamin D dose. Dotted line indicates the upper reference range of 10.2 mg/dL. Twenty episodes of hypercalcemia were observed in 14 women. No significant association between episodes of hypercalcemia and vitamin D dose was observed.

Supplemental Digital Content 1.

Baseline characteristics are shown in Table 1 . No significant differences between the eight groups were observed at baseline. The mean (SD) baseline serum 25(OH)D level was 15.6 ng/mL (39 nmol/L), which increased to 45 ng/mL (112 nmol/L) with the highest dose of vitamin D (4,800 IU/d). The mean (SD) daily dietary calcium intake was 685 (259) mg at baseline and 671 mg at 12 months. The mean calcium supplement was approximately 600 mg, and the mean total calcium intake at 12 months was 1,280 mg/day. The mean (SD) vitamin D intake was 114 (69) IU/day.

The mean (SD) overall adherence for 12 months was 94% (10%) for vitamin D3 and 91% (18%) for calcium.

Serum Calcium Results

The mean (SD) baseline serum calcium level was 9.47 mg/dL (2.36 nmol/L), which increased to 9.52 mg/dL (2.38 nmol/L) at 12 months ( P = 0.13). Twenty episodes of hypercalcemia (>10.2 mg/dL) were observed in 14 women (Figure 1, Table 2 ). The serum calcium level peaked at 3 months and slightly declined thereafter but never reached the baseline level. There was no significant association between episodes of hypercalcemia and vitamin D dose or serum 25(OH)D level. Hypercalcemia was transient and did not lead to reduction of calcium or vitamin D doses. In retrospect, two women may have had mild primary hyperparathyroidism. They had five episodes of hypercalcemia that had been excluded from the abovementioned analysis of hypercalcemia.

24-Hour Urine Calcium Results

The mean (SD) baseline 24-hour urine calcium level was 141 mg (3.5 mmol), which increased to 185 mg (4.6 mmol) at 12 months (P < 0.0001); changes at each time point are shown in Figure 2. Eighty-eight episodes of hypercalciuria (>300 mg) were observed in 48 women (Figure 3). When we excluded overcollection of 24-hour urine based on an individual's mean 24-hour urine creatinine level, the episodes of hypercalciuria were reduced from 88 to 81, and the number of women were reduced from 48 to 47 ( Table 2 ). These 81 episodes of hypercalciuria (>300 mg/d [7.7mmol]) occurred in 32% of women. Hypercalciuria occurred once in half of women and more than once in the other half of women; supplements were reduced in five women. The two women who may have had mild primary hyperparathyroidism did not have any episodes of hypercalciuria. There was no significant association between episodes of hypercalciuria and vitamin D dose or serum 25(OH)D level. Women who developed hypercalciuria or hypercalcemia were not the same people, and only two women developed both, as shown in Figure 4. All data were used to plot the graphs.

Figure 2.

Change in mean (95% CI) 24-hour urine calcium across time, adjusted for dose and estimated by a mixed-effects model.
*Significantly different from baseline (P < 0.0001).

Figure 3.

Incidence of hypercalciuria and vitamin D dose. Dotted line indicates an upper reference range of 300 mg. Eighty-one episodes of hypercalciuria (>300 mg) were observed in 47 women (32%) but were unrelated to vitamin D dose.

Figure 4.

All serum calcium and 24-hour urine calcium levels during the study. The upper reference range for serum calcium and 24-hour urine calcium is indicated by dotted lines. Only two women had both hypercalcemia and hypercalciuria; most women had either one or the other.

In 19 women, 24-hour urine calcium level increased to more than 400 mg (10 mmol); after repeated testing, urine calcium level decreased to below 300 mg (7.5 mmol) in 14 women and remained high in 5 women. Of these five women, calcium was discontinued in two women, and follow-up urine calcium level was less than 300 mg/day (7.5 mmol). However, hypercalciuria continued in two women, and both calcium and vitamin D were discontinued permanently. One woman refused follow-up testing. We found marginal associations between baseline 24-hour urine calcium and calcium intake estimated from the 7-day food diary (P < 0.074) and mean caffeine intake (P < 0.077).

Mixed-Effects Model

Log10 transformation of urine calcium improved the fit of the model. Model fit was examined with various residual plots and Akaike information criterion values, and the models were determined to fit well. In fitting the model, 161 women had urine calcium at baseline, 154 women at 3 months, 148 women at 6 months, 146 women at 9 months, and 146 women at 12 months. One hundred sixty-three women had serum calcium at baseline, 154 women at 3 months, 149 women at 6 months, 148 women at 9 months, and 147 women at 12 months. Dose-time interaction was not significant for urine calcium (P = 0.20) or serum calcium models (P = 0.17). Baseline levels of 24-hour urine calcium were significantly lower than the levels at all other time points (P < 0.0001), but none of the other time points differed significantly from each other (Figure 2). Urine calcium levels did not change significantly with vitamin D dose. There was a significant increase in serum calcium level from baseline to 6 months (P = 0.029). However, none of the other pairwise comparisons were significant, including those between baseline and 12 months (P = 0.56). Serum calcium levels did not change significantly with vitamin D dose.

PTH Results

At baseline, six women had high serum PTH levels (>65 pg/mL). Serum PTH levels decreased to the reference range at 6 months in four women and remained high at 6 and 12 months in two women. In retrospect, these two women may have had primary hyperparathyroidism, and their data on serum and 24-hour urine calcium were not used. The mean PTH level decreased at the end of the study, whereas serum and 24-hour urine calcium levels increased.

Multivariate Model

Generalized logistic regression was used to fit a multivariate model that predicts which women will develop hypercalciuria (300-400 and >400 mg). We considered baseline predictors that were significant in univariate comparisons, namely, age and urine calcium level at baseline. From the model, age and urine calcium level at baseline are significant predictors of hypercalciuria. A 1-year increase in age results in a 23% decrease in the risk of developing 24-hour urine calcium levels higher than 400 mg (10 mmol; odds ratio [OR], 0.77; 95% CI, 0.67-0.90; P = 0.0007). The odds of having a urine calcium level of 300 to 400 mg (7.5-10 mmol) versus a urine calcium level lower than 300 mg (7.5 mmol) did not differ significantly with age (P =0.16). For every 1-mg increase in baseline 24-hour urine calcium level, there was a 3% increase in the risk of having a urine calcium level higher than 400 mg (10 mmol; OR, 1.03; 95% CI, 1.02-1.04) and a 2% increase in the risk of having a urine calcium level between 300 and 400 mg (7.5-10 mmol; OR, 1.02; 95% CI, 1.01-1.03) compared with the reference group (urine calcium <300 mg [7.5 mmol]; P < 0.0001).

Prediction of Hypercalciuria Using ROC Curves for Urine Calcium at Baseline

We used ROC curves to determine a cutpoint of urine calcium at baseline that would predict a urine calcium level of 300 mg or higher during the course of the treatment. The ROC curve gave optimal sensitivity (87.2%) and specificity (70.1%), with an area under the curve of 0.85 (95% CI, 0.78-0.92) for a baseline urine calcium cutpoint higher than 132 mg ( Table 3 ).

We used a multivariate logistic regression model to predict a urine calcium level of 300 mg or higher during the course of the study, using the cutpoint determined for baseline urine calcium and age. A woman with a baseline urine calcium level higher than 132 mg has a 15 times greater risk of having a urine calcium level of 300 mg or higher during the course of the study than a woman with a baseline urine calcium level of 132 mg or less (OR, 15.3; 95% CI, 5.44-43.01; P < 0.0001). A 1-year increase in age results in a 10% decrease (or 0.90 times) in the risk of developing a urine calcium level of 300 mg or higher (OR, 0.90; 95% CI, 0.84-0.97; P = 0.0044).

Other cutpoints were explored for baseline urine calcium level using a multivariate logistic regression model ( Table 3 ). If the cutpoint was 180 mg or higher for baseline urine calcium level, there was a 20 times greater risk of having a urine calcium level of 300 mg or higher during the course of the study compared with a woman with a baseline urine calcium level lower than 180 mg (OR, 18.1; 95% CI, 6.7-49.2; P <0.0001).

Safety and Adverse Events

Eleven serious adverse events were observed in 11 women. Three of these adverse events were rated as severe, including diverticulitis, congestive heart failure, and tibia-fibula fracture; seven of these adverse events were described as moderate,including angina, syncope, hip surgical operation, radial fracture, chronic obstructive pulmonary disease exacerbation, stroke, and partial thyroidectomy for a nodule (none was attributed to vitamin D use). There was a small but significant increase of 0.068 mg/dL (95% CI, 0.041-0.096; P < 0.0001) in serum creatinine. One woman had an increase of 0.4 mg/dL, 4 women had an increase of 0.3 mg/dL, and 12 women had an increase of 0.2 mg/dL during 1 year. No symptomatic episodes of kidney stones were observed.

processing....