Risk Factors for Kidney Stones in Older Women in the Southern United States

Risk Factors for Kidney Stones in Older Women in the Southern United States W. DALLAS HALL, MD; MARY PETTINGER, MS; AL OBERMAN, MD, MPH; NELSON B. WATTS, MD; KAREN C. JOHNSON, MD, MPH; ELECTRA D. PASKETT, PHD; MARIAN C. LIMACHER, MD; JENNIFER HAYS, PHD

ABSTRACT: Background: The occurrence of kidney stones is disproportionate in the southern region of the United States. Risk factors for the occurrence of kidney stones in this geographic area have not been reported previously. Methods: The Women’s Health Initiative (WHI) is an ongoing multicenter clinical investigation of strategies for the prevention of common causes of morbidity and mortality among postmenopausal women. A case-control ancillary study was conducted on 27,410 (white or black) women enrolled in the 9 southern WHI clinical centers. There were 1,179 cases (4.3%) of kidney stones at the baseline evaluation. Risk factors for stone formation were assessed in cases versus age- and race-matched control subjects. Results: Risk factors (univariate) included low dietary potassium (2,404 versus 2,500 mg/day, P ⫽ 0.006), magnesium (243 versus 253

mg/day, P ⫽ 0.003) and oxalate (330 versus 345 mg/ day, P ⫽ 0.02) intake, as well as increased body mass index (28.5 versus 27.7 kg/m2, P ⫽ 0.001) and a history of hypertension (42% versus 34%, P ⫽ 0.001). A slightly lower dietary calcium intake (683 versus 711 mg/day, P ⫽ 0.04) was noted in case subjects versus control subjects, but interpretation was confounded by the study of prevalent rather than incident cases. Supplemental calcium intake ⬎500 mg/day was inversely associated with stone occurrence. Conclusion: Multivariate risk factors for the occurrence of kidney stones in postmenopausal women include a history of hypertension, a low dietary intake of magnesium, and low use of calcium supplements. KEY INDEXING TERMS: Kidney stones; Women; Diet; Hypertension. [Am J Med Sci 2001; 322(1):12–18.]

A

dietary intake of oxalate,8 sodium,9 sucrose,10 or animal protein.11 Inverse associations have been noted for high dietary intake of calcium,6,7 potassium,12,13 and magnesium.12,13 Other reported risk factors for stone formation include white race, hypertension, low alcohol intake, and such climatologic factors as increased heat and sunlight.1,5,14 –20 The Women’s Health Initiative (WHI) is an ongoing multicenter clinical investigation of strategies for the prevention of the common causes of morbidity and mortality among postmenopausal women.21 Between 1992 and 1998, WHI enrolled 161,859 postmenopausal women. Of these, 32,072 resided in the southern United States; the 27,410 who were black or white form the basis of this report. The kidney stone risk factors assessed in this study included dietary intake of calcium, oxalate, sodium, potassium, magnesium, and other nutrients. Also assessed were alcohol intake, supplemental calcium intake, history of hypertension, use of thiazide diuretics, and the average ambient temperature of the geographic location where the participant was enrolled.

disproportionate geographic occurrence of kidney stones is apparent in the southern United States,1,2 similar to the “stroke belt” in this same region.3 Figure 1 is an illustration of the regional prevalence of kidney stones according to data collected incidentally during a survey of 1,185,124 men and women who participated in a 1982 50-state Cancer Prevention Survey conducted by the National Cancer Institute.1,4 The overall incidence of kidney stones is about 3-fold higher in men than in women.5 The same is true in the southeastern United States.1 Reported dietary risk factors for the occurrence of stones include high supplemental calcium intake6,7 and high From the Emory University School of Medicine, Atlanta, Georgia (WDH, NBW); Fred Hutchinson Cancer Research Center, Seattle, Washington (MP); University of Alabama, Birmingham, Alabama (AO); University of Tennessee, Memphis, Tennessee (KCJ); Wake Forest University School of Medicine, Winston-Salem, North Carolina (EDP); , University of Florida, Gainesville, Florida (MCL); and University of Texas, Houston, Texas (JH). Submitted January 25, 2001; accepted March 28, 2001. This study was supported by NHLBI research contracts for the Women’s Health Initiative. Correspondence: W. Dallas Hall, M.D., Emeritus Professor of Medicine, 1100 Parker Place, Atlanta, GA 30324-5402 (E-mail: [email protected]).

12

Methods Study Population. All 27,410 study participants were postmenopausal women aged 50 to 79 years, enrolled into WHI July 2001 Volume 322 Number 1

Hall et al

computed. All P-values are two-sided. Statistical analyses and the matching of case subjects and control subjects were performed using SAS (SAS Institute, Cary, NC). The dietary intake data were calculated as both absolute values and as milligrams per 1000 kilocalories. Absolute values (and their levels of significance) are reported in the tables to allow comparison with other nutrient data and national guidelines. Analyses were performed using both the absolute and adjusted values. There were minor changes in the P values, but in no case were there any modifications in the interpretation or conclusions.

Results

Figure 1. Regional prevalence (%) of kidney stones in U.S. white women. [Adapted from Soucie JM, Thun MJ, Coates RJ, et al. Demographic and geographic variability of kidney stones in the United States. Kidney Int 1994;46:893–9. Copyright © 1994 Blackwell Science. Used with permission.]

between 1992 and 1998 at 1 of 9 southern WHI clinical centers or their satellite clinic (Atlanta, GA; Birmingham, AL; Chapel Hill, NC; Gainesville, FL; Houston, TX; Memphis, TN; Miami, FL; San Antonio, TX; Winston-Salem, NC). Cases (1,179) were defined as women who answered “yes” to a history of kidney or bladder stones as part of a baseline questionnaire. Control subjects (1 per case: 1,179) answered “no” to the same question. They were selected randomly and were matched for race, age ⫾ 5 years, and the WHI study component (ie, randomized clinical trial or observational study). Matching was performed using an algorithm that minimizes an overall distance measure that represents the total of absolute deviations for all matching criteria.22 Ethnic groups other than white or black were excluded because of the low absolute number of stone cases available for analysis. Potential participants (5.6%) were excluded if the baseline dietary variable data were incomplete. Data Collection. The Coordinating Center (Seattle, WA) trained certified specific staff (of each of the 9 clinical centers) in the method of data collection. These staff members then obtained all data at the time of baseline enrollment and written consent, before any assigned study interventions. A standardized questionnaire included demographic variables, previous medical diagnoses, alcohol and supplemental calcium intake, and over-thecounter medications. Dietary intake variables were calculated from a modified food frequency questionnaire.23 Body weight was measured using calibrated scales to the nearest half-pound, lightly clothed, without shoes. Height was measured to the nearest half-inch, using a stadiometer. Body mass index (BMI) was calculated as weight in kilograms divided by the height in meters squared. Overweight was defined as a BMI ⱖ25 kg/m2 and obesity as a BMI ⱖ30 kg/m2.24 Hypertension was defined as patient history or the current use of antihypertensive medications. Average annual ambient temperatures for the 9 clinical centers were obtained from the National Weather Service web site.25 Statistical Analysis. Univariate methods were used to compare the demographic, calcium supplement and medication use, and dietary intake variables between case subjects and control subjects. Statistical tests of categorical variables were performed using a ␹2 statistic. t Tests were used to test for differences in mean levels of continuous variables. Tests of the dietary intake variables were based on log-transformed values. A multivariate logistic regression model assessed the odds of the presence of kidney stones for each of the variables evaluated univariately, simultaneously adjusting for all other variables in the model. Adjustment was also made for clinical center. The significance of interactions was assessed using the improvement ␹2 statistic based on the difference in deviance between the nested models. For all odds ratios, 95% confidence intervals were

THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

Baseline Characteristics. The average age of the case subjects (63.5 ⫾ 7.1 years) and control subjects (63.5 ⫾ 7.1 years) was identical. Of the 2,358 participants, 88.5% were white and 11.5% were black. As anticipated from the matching, the mean age and ethnic distribution did not differ significantly between case subjects and control subjects. Control subjects, however, had somewhat higher levels of educational achievement (P ⫽ 0.02). Dietary and Supplemental Calcium Intake. The dietary calcium intake was slightly lower (P ⫽ 0.04) in case subjects (683 mg/day; 437 mg/1000 kcal) versus control subjects (711 mg/day; 458 mg/ 1000 kcal). Supplemental calcium intake was lower in case subjects (273 mg/day) versus control subjects (380 mg/day). Almost half of all participants (49% of case subjects and 42% of control subjects), however, were taking no calcium supplements. Among the case subjects and control subjects taking calcium supplements (⬎0 –100, ⬎100 –500, and ⬎500 mg/ day), there was a lower prevalence of stones in women taking more than 500 mg/day. The average dietary oxalate intake was lower (P ⫽ 0.02) in case subjects (330 mg/day; 220 mg/1000 kcal) versus control subjects (345 mg/day; 231 mg/1000 kcal) (Tables 1 and 2). Dietary Sodium, Potassium, and Magnesium Intake. Dietary sodium intake did not differ (P ⫽ 0.91) between case subjects (2608 mg/day) and control subjects (2611 mg/day) (Table 1). There was no effort to quantify or include the intake of table salt. Dietary potassium intake was lower (P ⫽ 0.006) in case subjects (2404 mg/day) versus control subjects (2500 mg/day). Dietary magnesium intake was also lower (P ⫽ 0.003) in case subjects (243 mg/day) versus control subjects (253 mg/day). Alcohol Intake. Alcohol intake was reported more often (P ⫽ 0.001) in control subjects (49%) than case subjects (42%). Among women drinking any alcohol, however, there was no difference (P ⫽ 0.70) in the amount consumed per day between case subjects and control subjects (Tables 1 and 2). Hypertension and Thiazide Diuretic Use. A history of hypertension was present in 38% of all participants, more so in case subjects (485/1170 ⫽ 41.5%) than control subjects (402/1169 ⫽ 34.4%) (P ⫽ 0.001). The majority of hypertensive case subjects (74%) and control subjects (79%) were receiving an13

Kidney Stones in Women

Table 1. Kidney Stones in Older Southern U.S. Women: Descriptive Analyses Controls (N ⫽ 1179) a

Variable

Dietary (per day) Energy (kcal) Calcium (mg) Sodium (mg) Potassium (mg) Magnesium (mg) Protein (g) Animal protein (g) Oxalic acid (mg) Sucrose (g) Fruits and vegetables (servings/day) Other Calcium supplements (mg/day) Alcohol intake (g/day) BMI (kg/m2) Age (years)

Cases (N ⫽ 1179)

Mean

SD

Mean

SD

Pb

1563 711 2611 2500 253 66 46 345 36 3.9

585.2 385.1 1044.0 928.7 93.5 27.5 22.4 165.8 18.6 2.1

1573 683 2608 2404 243 65 46 330 36 3.8

590.0 377.1 1032.9 928.1 95.4 26.9 21.4 161.2 19.1 2.1

0.82 0.04 0.91 0.006 0.003 0.41 0.59 0.02 0.37 0.04

380 4.0 27.7 63.5

534.7 8.8 6.1 7.1

273 3.4 28.5 63.5

444.2 8.4 6.1 7.1

0.001 1.00

a

Two case subjects and three control subjects had missing dietary data; 10 case subjects and 10 control subjects had missing BMI values. t Tests were used to test for differences in mean levels between cases and controls. Tests of the nutrients were based on log-transformed values. No test was performed for differences in calcium supplements or alcohol intake because so many of the observations had zero values (see categorical data analyses, Table 2).

b

tihypertensive treatment. The overall use of thiazide diuretics was relatively low and did not differ between case subjects (7.3%) and control subjects (6.5%) (P ⫽ 0.47) (Table 2). The dietary sodium intake of the 887 women with a history of hypertension (2624 mg/day) did not differ (P ⫽ 0.75) from that of the 1452 normotensives (2599 mg/day) (data not shown). Body Mass Index. Average body height was similar for case subjects (162.1 cm) and control subjects (162.3 cm) (Tables 1 and 2). Average body weight was greater (P ⫽ 0.001) for case subjects (75.6 kg) versus control subjects (73.2 kg). BMI also was greater (P ⫽ 0.001) in case subjects (28.5 kg/m2) versus control subjects (27.7 kg/m2). Overweight (ie, BMI ⱖ25 kg/m2) was present in 806 of 1169 case subjects (68.9%) versus 719 of 1169 control subjects (61.5%) (P ⫽ 0.001). Obesity (ie, BMI of ⱖ30 kg/m2) was present in 387 of 1169 case subjects (33.1%) versus 329 of 1169 control subjects (28.1%) (P ⫽ 0.004). Among all participants, BMI was greater in those with a history of hypertension (29.9 ⫾ 6.3 kg/m2) compared with normotensives (27.0 ⫾ 5.7 kg/m2) (P ⫽ 0.0001) (data not shown). Ambient Temperature. Table 3 shows the average annual ambient temperature relative to the prevalence of kidney stone case subjects in each of the cities represented. The average temperature was 64.3°F for case subjects and 64.9°F for control subjects. There was no association between annual ambient temperature and stone occurrence. The average annual temperature, however, varied only between 59°F and 76°F within the southern region. 14

Logistic Regression. A multivariate analysis assessed the factors related to the occurrence of kidney stones. History of hypertension, a low dietary magnesium intake, and low use of calcium supplements were associated with kidney stones (Tables 4 and 5). Terms for the interaction of supplemental calcium intake with alcohol intake, BMI, and a history of hypertension were added separately to the multivariate logistic regression model to assess the effect on the occurrence of kidney stones. None of these interactions was statistically significant (P ⬎ 0.50 for each). Discussion One strength of this study is the unusually large sample of women (1179) with a history of kidney stones. A prospective study of incident (rather than prevalent) cases would have been even more desirable, but these data will not be available in the Women’s Health Study (WHI) until approximately 2005. The large number of WHI enrollees allowed for complete matching (1:1) of the case subjects with control subjects of the same race, age ⫾ 5 years, clinical center, and WHI study component. In addition, the detailed dietary and other data were collected and analyzed in an identical manner for case subjects and control subjects, and by the same trained research staff team. The study sample included only women; conclusions about men can be reached only by inference. Many of the same risk factors (ie, location, climate, diet, blood pressure, etc), however, would be shared by both genders in this region. July 2001 Volume 322 Number 1

Hall et al

Table 2. Kidney Stones in Older Southern U.S. Women: Categorical Data Analyses Control Variable Hypertension historyb No Yes Thiazide use No Yes Calcium supplements No Yes Calcium supplements (mg/day) ⬎0–100 ⬎100–500 ⬎500 Alcohol intake No Yes Alcohol intake (g/day) ⬎0–2 ⬎2–4 ⬎4 Current smoker No Yes BMI (kg/m2)c ⬍19 19–24.9 25–29.9 30–39.9 ⱖ40

Case

N

%

N

%

767 402

65.6 34.4

685 485

58.5 41.5

1102 77

93.5 6.5

1093 86

92.7 7.3

496 683

42.1 57.9

572 607

48.5 51.5

43 278 362

6.3 40.7 53.0

58 318 231

9.6 52.4 38.1

597 582

50.6 49.4

681 498

57.8 42.2

240 70 272

41.2 12.0 46.7

216 62 220

43.4 12.5 44.2

1082 84

92.8 7.2

1069 96

91.8 8.2

25 425 390 281 48

2.1 36.4 33.4 24.0 4.1

17 346 419 330 57

1.5 29.6 35.8 28.2 4.9

Pa 0.001 0.47 0.002 0.001

0.001 0.70

0.35 0.004

␹2 test. Nine case subjects and 10 control subjects had missing data. c Ten case subjects and 10 control subjects had missing data. a b

The Health Professionals Follow-Up Study reported in 1993 that a high dietary calcium intake was associated with a lower risk of kidney stones (505 cases) in a cohort of 45,619 men aged 40 to 75 years.6 The result was opposite to the previous concept that a higher calcium intake might increase the risk of kidney stones. Four years later, the Nurses’ Health Study also reported that higher dietary calcium intake was associated with a lower risk of kidney stones (864 cases) in a cohort of 91,731 women aged 34 to 59 years (residing in 11 different states).7 In both studies, the protective effect of a higher dietary calcium intake was conceptually related to a decreased absorption and excretion of oxalate. It was not feasible at that time to accurately estimate total dietary oxalate intake. The present study demonstrated a relatively low level of dietary calcium intake in both case subjects (683 mg/day; 437 mg/1000 kcal) and control subjects (711 mg/day; 458 mg/1000 kcal). This level of dietary calcium intake in women in the southern United States is markedly below that reported in the previous 2 studies of health professionals. In our study, the dietary calcium intake was significantly lower in THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

case subjects than control subjects (P ⫽ 0.04), but interpretation of these data is muted by the use of prevalent cases. The use of prevalent rather than incident cases could introduce selection bias in the history of dietary or supplemental calcium intake. Women with kidney stones may have been (inappropriately) advised by their physician or decided on their own to reduce their calcium intake after the diagnosis. In general, however, our results are compatible with the findings of the Health Professionals Follow-Up Study and Nurses’ Health Study that higher dietary calcium intake is not associated with a greater risk of kidney stones. In contrast to the findings on dietary calcium, the Nurses’ Health Study reported a modest (relative risk, 1.2) direct association between supplemental calcium intake and the risk of kidney stones.7 Calcium supplements were used by only 55% of the women in our study. The use of calcium supplements was more likely in control subjects than case subjects, and among the women taking supplements, more control women were taking the higher doses. The odds ratio for kidney stones was 0.62 (95% confidence limits, 0.50 to 0.77) in 593 women 15

Kidney Stones in Women

Table 3. Kidney Stone Prevalence and Average Ambient Temperature of Southern WHI Clinical Centers

Location

Average Annual Temperature (°F)

Stone Cases (N)

Total Women (N)

Stone Prevalence (%)

Miami, FL Gainesville, FL Houston, TX San Antonio, TX Jacksonville, FLb Birmingham, AL Memphis, TN Memphis, TNb Atlanta, GA Winston-Salem, NCc Winston-Salem, NCb,c Chapel Hill, NCc TOTAL or AVERAGE

75.9 68.4 69.4 68.6 68.4 62.0 62.0 62.0 61.2 59.0 59.0 59.0 64.6

65 112 98 57 90 156 123 33 147 109 48 141 1179

1746a 2266 2869a 1764a 2090 3447 2662 837 3505 2008 945 3311 27410

3.72 4.94 3.42 3.23 4.31 4.53 4.69 3.94 4.19 5.43 5.08 4.26 4.30

a

Hispanic participants not included Satellite clinical center c Represented by Raleigh, NC, weather data b

taking more than 500 mg/day of supplemental calcium. Only recently has it been possible to accurately estimate the dietary intake of oxalate from the full range of foods in the food frequency questionnaire. Oxalate-rich foods (eg, collards, greens, grits, and

tea) are popular in the South, but our data show a significantly lower dietary oxalate intake in case subjects compared with control subjects. Urinary oxalate excretion, however, is more a function of oxalate absorption than of dietary oxalate intake.7,26 An increase in calcium intake inhibits oxalate ab-

Table 4. Kidney Stones in Older Southern U.S. Women: Logistic Regression Results Variable

Odds Ratioa

95% Confidence Interval

1.19 1.07 0.91 1.12 0.67 1.99 0.55 0.96 0.97

0.90–1.58 0.91–1.27 0.70–1.19 0.79–1.59 0.48–0.95 0.88–4.54 0.31–0.99 0.81–1.14 0.85–1.10

0.23 0.42 0.48 0.53 0.02 0.10 0.05 0.62 0.61

1.00 1.24

1.03–1.49

0.02

1.00 0.86

0.61–1.21

0.37

1.00 0.88 0.92 0.84

0.70–1.10 0.63–1.34 0.66–1.06

0.27 0.67 0.14

1.00 1.33 1.04 0.62

0.86–2.04 0.85–1.29 0.50–0.77

0.20 0.69 0.0001

1.00 1.18 1.44 1.49

0.62–2.24 0.76–2.74 0.78–2.84

0.61 0.27 0.23

Dietary (per day) Energy (kcal) Calcium (mg) Sodium (mg) Potassium (mg) Magnesium (mg) Protein (g) Animal protein (g) Oxalic acid (mg) Sucrose (g) Hypertension History No Yes Thiazide Use No Yes Alcohol Use (g/day) 0 ⬎0–2 ⬎2–4 ⬎4 Calcium Supplements (mg/day) 0 ⬎0–100 ⬎100–500 ⬎500 BMI (kg/m2) ⬍19 19–24.9 25–29.9 ⱖ30 a

P

Odds ratio associated with an upward shift of 1 SD.

16

July 2001 Volume 322 Number 1

Hall et al

Table 5. Risk Factors for the Occurrence of Kidney Stones in Older Women Residing in the Southern United States Multivariate History of hypertension Low dietary intake of magnesium Supplemental calcium intake ⬍ 500 mg/day Univariate Increased body mass index Low dietary intake of calcium Low dietary intake of oxalate Low dietary intake of potassium

sorption and excretion, whereas a decrease in calcium intake enhances oxalate absorption and excretion. High dietary intake of sodium and low dietary intake of potassium or magnesium may increase the risk of nephrolithiasis, possibly through an increase in urinary calcium excretion.9,12 In contrast to some previous reports,7,9 we found no greater dietary sodium intake in case subjects than control subjects. Dietary potassium and magnesium intakes were lower in case subjects than control subjects. The average dietary intake of potassium, however, was relatively low (2404 to 2500 mg/day) in southern U.S. women. We found no significant relationship between dietary sucrose or protein intake and stone occurrence. A recent report found a direct association of kidney stones with an increased intake of grapefruit juice.27 Grapefruit juice intake was not quantified specifically in the current study. Alcohol intake has been reported to correlate inversely with stone occurrence.5,17 Kidney stones were reported more often among our 1278 postmenopausal southern women who were teetotalers (54%), but alcohol use was not a significant independent risk factor. Moreover, there was no evidence for any trend between stone occurrence and the amount of alcohol consumed in users (Tables 2 and 4). A possible explanation for the notable geographic disparity for the prevalence of kidney stones is the dissimilar climates. Warmer climates, especially during the summer months, seem to predispose to kidney stones. In the United States, there is a northsouth gradient of kidney stones that correlates well with differences in ambient temperatures and a sunlight index, which is 2 times higher in the southeast compared with the northeast. Other geographical regions throughout the world, such as the desert regions, have a high incidence of stones.18 In the cross-sectional Second Cancer Prevention Survey among more than 1 million persons, ambient temperature and sunlight were independently associated with an increased prevalence of stones.1,4 Controlling ambient temperature, beverage intake, and sunlight index markedly reduced the geographic variability noted in the U.S. Other factors did not seem to explain the variability in kidney stones. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES

Though the study was a prevalence study, it is unlikely that persons would move to a warmer climate because of kidney stones. An increased BMI has not been previously reported as a risk factor for kidney stones. Our data, however, show a much higher prevalence of both overweight and obesity with kidney stones. The mechanism for this association is unclear, although it is partly accounted for by the overweight profile of patients with a history of hypertension. The recently recognized association between hypertension and kidney stones in men2,15,16 was confirmed by this study in women. The association was independent of related variables such as age, BMI, or the use of thiazide diuretics. Our data do not allow us to ascertain whether the kidney stones preceded the hypertension or vice versa. Nonetheless, clinicians must now be more aware of the cooccurrence of hypertension and nephrolithiasis. Hypertension is the major risk factor for stroke, and it may be of more than incidental interest that the “stroke belt” and the “stone belt” include primarily southeastern states. In the logistic regression results, the 3 independent risk factors for the occurrence of kidney stones in southern women were a history of hypertension, a low dietary magnesium intake, and a low use of calcium supplements. References 1. Soucie JM, Thun MJ, Coates RJ, et al. Demographic and geographic variability of kidney stones in the United States. Kidney Int 1994;46:893–9. 2. Soucie JM, Coates RJ, McClellan W, et al. Relation between geographic variability in kidney stones prevalence and risk factors for stones. Am J Epidemiol 1996;143:487–95. 3. Hall WD, Ferrario CM, Moore MA, et al. Hypertensionrelated morbidity and mortality in the southeastern United States. Am J Med Sci 1997;313:195–209. 4. Stellman SD, Garfinkel L. Smoking habits and tar levels in a new American Cancer Society prospective study of 1.2 million men and women. J Natl Cancer Inst 1986;76:1057– 63. 5. Johnson CM, Wilson DM, O’Fallon WM, et al. Renal stone epidemiology: A 25-year study in Rochester, Minnesota. Kidney Int 1979;16:624 –31. 6. Curhan GC, Willett WC, Rimm EB, et al. A prospective study of dietary calcium and other nutrients and the risk of symptomatic kidney stones. N Engl J Med 1993;328:833– 8. 7. Curhan GC, Willett WC, Speizer FE, et al. Comparison of dietary calcium with supplemental calcium and other nutrients as factors affecting the risk for kidney stones in women. Ann Intern Med 1997;126:497–504. 8. Borsatti A. Calcium oxalate nephrolithiasis: defective oxalate transport. Kidney Int 1991;39:1283–98. 9. Cirillo M, Laurenzi M, Panarelli W, et al. Urinary sodium to potassium ratio and urinary stone disease. The Gubbio Population Study Research Group. Kidney Int 1994;46: 1133–9. 10. Lemann J Jr, Piering WF, Lennon EJ. Possible role of carbohydrate-induced calciuria in calcium oxalate kidneystone formation. N Engl J Med 1969;280:232–7.

17

Kidney Stones in Women

11. Breslau LA, Brinkley L, Hill KD, et al. Relation of animal protein-rich diet to kidney stone formation and calcium metabolism. J Clin Endocrinol Metab 1988;66:140 – 6. 12. Lemann J Jr, Pleuss JA, Gray RW, et al. Potassium administration reduces and potassium deprivation increases calcium excretion in healthy adults. Kidney Int 1991;39:973– 83. 13. Kesteloot H, Joosens JV. Relationship of dietary sodium, potassium, calcium, and magnesium with blood pressure: interuniversity research on nutrition and health. Hypertension 1988;12:594 –9. 14. Johansson G, Backman U, Danielson BG, et al. Biochemical and clinical effects of the prophylactic treatment of renal calcium stones with magnesium hydroxide. J Urol 1980;124: 770 – 4. 15. Cappuccio FP, Strazzullo P, Mancini M. Kidney stones and hypertension: Population based study of an independent clinical association. BMJ 1990;300:1234-6. 16. Madore F, Stampfer MJ, Rimm EB, et al. Nephrolithiasis and risk of hypertension. Am J Hypertens 1988;11:46 – 53. 17. Krieger JN, Kronmal RA, Coxon V, et al. Dietary and behavioral risk factors for urolithiasis: Potential implications for prevention. Am J Kidney Dis 1996;28:195–201. 18. Pierce LW, Bloom B. Observations on urolithiasis among American troops in a desert area. J Urol 1945;54:466 –70. 19. Parry ES, Lister IS. Sunlight and hypercalciuria. Lancet 1975;1:1063–5.

18

20. Rodgers AL, Greyling KG, Irving RA, et al. Crystalluria in marathon runners. II. Ultra-marathon males and females. Urol Res 1988;16:89 –93. 21. Women’s Health Initiative Study Group. Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials 1998;19:61–109. 22. Bergstralh EJ, Kosanke JL. Computerized matching of cases and controls. Technical Report #56. Rochester (MN): Department of Health Sciences Research, Mayo Clinic; 1995. 23. Patterson RE, Kristal AR, Tinker LF, et al. Measurement characteristics of the Women’s Health Initiative food frequency questionnaire. Ann Epidemiol 1999;9:178 – 87. 24. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults. The Evidence Report. NHLBI Obesity Education Initiative Expert Panel. NIH Publication No. 98-4083. Bethesda (MD): National Heart, Lung and Blood Institute; 1998. 25. NWS climate table. National Weather Service Web Site. Available from: URL: http://www.nws.noaa.gov/climatex.html. Accessed October 1999. 26. Massey LK, Roman-Smith H, Sutton RA. Effect of dietary oxalate and calcium on urinary oxalate and risk of formation of calcium oxalate kidney stones. J Am Diet Assoc 1993;93:901– 6. 27. Curhan GC, Willett WC, Speizer FE, et al. Beverage use and risk for kidney stones in women. Ann Intern Med 1998; 128:534 – 40.

July 2001 Volume 322 Number 1