Family History and Risk of Kidney Stones

GARY C. CURHAN,*�* WALTER C. WILLETT,*� ERIC B. RIMM,*t and

MEIR J. STAMPFER*t�From the Departments of *Nutrition and tEpidemiology, Harvard School of Public Health, Boston,

Massachusetts; �Channing Laboratory, Department of Medicine, Brigham and Women ‘s Hospital, Boston,

Massachusetts; �Renal Unit, Department of Medicine, Massachusetts General Hospital, Boston,

Massachusetts; and Harvard Medical School, Boston, Massachusetts.

Abstract. Kidney stones develop more frequently in individuals

with a family history of kidney stones than in those without a

family history; however, little information is available regard-

ing whether the increased risk is attributable to genetic factors,

environmental exposures, or some combination. In this report,

the relation between family history and risk of kidney stone

formation was studied in a cohort of 37,999 male participants

in the Health Professionals Follow-up Study. Information on

family history, kidney stone formation, and other exposures of

interest, including dietary intake, was obtained by mailed ques-

tiommaires. A family history of kidney stones was much more

common in men with a personal history of stones at baseline in

1986 than in those without a history of stones (age-adjusted

prevalence odds ratio, 3. 16; 95% confidence interval [CI], 2.90

to 3.45). During 8 yr of follow-up, 795 incident cases of stones

were documented. After adjusting for a variety of risk factors,

the relative risk of incident stone formation in men with a

positive family history, compared with those without, was 2.57

(95% CI, 2. 19 to 3.02). Family history did not modify the

inverse association between dietary calcium intake and the risk

of stone formation. These results suggest that a family history

of kidney stones substantially increases the risk of stone for-

mation. In addition, these data suggest that dietary calcium

restriction may increase the risk of stone formation, even

among individuals with a family history of kidney stones.

(J Am Soc Nephrol 8: 1568-1573, 1997)

Kidney stones develop more frequently in individuals with a

family history of kidney stones than in those without a family

history; however, little information is available regarding

whether the increased risk is attributable to genetic factors,

environmental exposures or some combination. A positive

family history of stones has been reported in 16% (1) to 37%

(2) of patients who have formed a kidney stone, compared with

4% to 22% (1 ,3,4) in healthy control subjects. Proposed inher-

ited traits predisposing to stone formation include hypercalci-

uria (5,6), defective oxalate transport (7), incomplete renal

tubular acidosis (8), and increased uric acid production. Envi-

rommental exposures, such as diet, are thought to play an

important role in stone formation (9); however, scant data exist

on the potential interaction between family history of kidney

stones and environmental influences.

To examine the influence of family history on kidney stone

formation, we studied this relation in a cohort of 37,999 male

participants in the Health Professionals Follow-up Study. We

also examined the potential interaction between family history

and dietary factors on risk of stone formation. In addition, we

compared the results from 24-h urine collections in men with

and without a family history of kidney stones.

Received November 19, 1996. Accepted March 14, 1997.

Correspondence to Dr. Gary C. Curhan, Department of Nutrition, Harvard

School of Public Health, 665 Huntington Avenue. Boston. MA 021 15.

1046-6673/08010- 1568$03.00/0

Journal of the American Society of Nephrology

Copyright U 1997 by the American Society of Nephrology

Materials and MethodsStudy Population

The Health Professionals Follow-up Study is a longitudinal studyof diet and disease among 5 1 ,529 male dentists, optometrists, osteo-

paths, pharmacists, podiatrists, and veterinarians who were 40 to 75 yr

of age in 1986. The participants returned a mailed questionnaire in

1986 concerning diet, medical history, and medications. The cohort is

followed-up 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 37,999 men who answered the 1994

long-form questionnaire, which included a question about family

history of kidney stones.

Assessment of Family History

On the 1994 questionnaire, all participants were asked if eitherparent or any of their siblings had ever had a kidney stone. The

relatives of the participants were not contacted by the investigators.

Assessment of Diet

Diet was assessed in 1986 and 1990 using a semiquantitative

food-frequency questionnaire that inquired about the average use of

13 1 foods and beverages during the previous year. Nutrient intake was

computed from the reported frequency of consumption of each spec-

ified unit of food or beverage and from published data on the nutrient

content of the specified portions ( 10). Information was also collected

on supplemental vitamins and minerals, including calcium (such as

calcium carbonate), either alone or in multivitamin preparations. We

have previously reported on the reproducibility and validity of this

dietary questionnaire in this cohort (1 1).

Nutrient values were adjusted for total energy intake using a

regression model with total caloric intake as the independent variable

Family History and Kidney Stones 1569

and absolute nutrient intake as the dependent variable (12). Because

total energy intake for an individual tends to be fixed in a narrow

range, changes in nutrient intake must be made primarily by altering

the 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 any variation introduced by underreport-

ing or overreporting of intake on the food frequency questionnaire,

thus improving the accuracy of nutrient measurement ( I 1).

24-H Urine Collections

In 1994, we invited participants with incident stones confirmed bysupplementary questionnaire and a random sample of control subjects

to provide a 24-h urine collection. Participants were ineligible for the

urine collection study if they were older than 70 yr of age, had a

history of cardiovascular disease or cancer, had their first stone prior

to 1988, or did not respond to the 1994 biennial questionnaire. We

chose not to request urine from men who had stones in the very distant

past and thus excluded men from the urine collection who had their

incident stone between 1986 and 1988. The men were asked to collect

their urine for measurements of calcium, citrate, oxalate, potassium,

and other factors, using a collection system developed by Mission

Pharmacal (San Antonio, TX), which can be mailed to participants

and then returned by mail to Mission for analysis. The men were sent

a 4-L jug with a lithium-impregnated sponge attached to the bottom

and were asked to collect all of their urine in the jug during a 24-h

period. At the completion of the collection, samples were poured into

two small vials, which were provided. These vials were placed in a

self-addressed stamped mailer and returned to Mission. The labora-

tory measurements were performed by technicians at Mission Phar-

macal who were blinded to the individuals’ exposures and case status.

Total volume is calculated from the concentration of lithium in thesamples. In 1994, 79% of eligible men with confirmed stones and

59% of control subjects agreed to participate, and the collections were

completed by 302 (77%) and 95 (81%), respectively. Sixteen men

with incomplete collections, determined by total urinary creatinine

excretion of less than 1000 mg/d or total urine volume of less than 400

mLfd, and I 2 with missing family history information were excluded

from the analysis using the urine data, leaving a total of 369 samples.

Based on supplementary information provided by the case subjects,

less than I 0% of the case subjects who performed the urine collection

were being treated with medication to prevent stone recurrence.

Follow-Up and Case Ascertainment

We sent follow-up questionnaires to the entire cohort in 1988,

1990, 1992, and 1994, inquiring about kidney stones diagnosed sinceJanuary 1986. After up to six mailings for each follow-up period, theresponse rate averaged 94% for each mailing cycle.

Prevalent cases were considered those that occurred before 1986.

We considered incident cases those that occurred during the first 8 yrof follow-up, between the return of the 1986 baseline questionnaire

and lanuary 3 1, 1994. If a kidney stone was reported on a follow-up

questionnaire, then a supplementary form was mailed to confirm the

self-report and to ascertain the date of occurrence and symptoms. The

response rate to the supplementary questionnaire for each 2-yr period

was over 92%. The primary end point was an incident kidney stone

accompanied by pain or hematuria. To confirm the validity of the

self-report, we obtained the medical records from a random sample of

60 of the confirmed cases. The records confirmed the diagnosis of akidney stone in 97% of the cases; the remaining 3% were bladder

stones.

Statistical Analyses

For each participant, the person-months of follow-up were counted

from the date of the return of the 1986 questionnaire to the date of a

kidney stone or death, or January 3 1 , 1994, whichever came first.

Information on exposures of interest from the 1986 questionnaire was

updated in 1990. We allocated person-months of follow-up according

to exposure status at the start of each follow-up period (e.g. , indicated

by the quintile of calcium intake and other variables) and calculated

incidence as the number of events divided by the person-time of

follow-up. If complete dietary information was missing at the start of

a time period, the subject was excluded for that time period.

The relative risk-the incidence rate in a particular category of

exposure divided by the corresponding rate in the comparison cate-

gory-was used as the measure of association. Age-adjusted relative

risks were calculated after stratification by 5-yr age categories. The

Mantel extension test was used to evaluate linear trends across cate-gories of nutrient intake. In addition, relative risks were adjusted

simultaneously for potentially confounding variables, using a propor-

tional-hazards model (1 3). The variables considered in these models

were age (5-yr categories), geographic region (seven categories), use

of thiazide diuretics (yes/no), alcohol intake (seven categories), body

mass index (six categories), quintiles of total fluid intake, supplemen-

tal calcium intake (none, I to 100 mg/d, 101 to 500 mg/d, and >500

mg/d) and quintiles of energy-adjusted dietary intake of calcium,

animal protein, sucrose, magnesium. sodium, potassium, and caffeine.

Separate logistic regression models were run after stratifying the menaccording to family history of kidney stones. Logistic models were

also run using the raw nutrient values without energy-adjustment. For

all relative risks, we calculated 95% confidence intervals (CI).

Categorical comparisons were performed using Fisher’s exact test,

and confidence intervals were calculated using the RBV variance

(StatXact, Cambridge, MA). Continuous variables were comparedusing the t test (SAS Institute, Cary, NC). All P values are two-tailed.

ResultsOf the 37,999 men providing information in 1994, a family

history of kidney stones was reported by 4873 (12.8%).

Prevalent Stones

In 1986, 2957 (7.8%) men had a personal history of kidney

stones. A family history of kidney stones was reported by

17.2% of men who had a kidney stone before 1986, compared

with 6.4% of men who had never had a stone (age-adjusted

odds ratio = 3. 16; 95% CI, 2.90 to 3.45) (Table 1).

Incident Stones

For the analyses of incident stones occurring between 1986

and 1994, we excluded men with a personal history of kidney

stones at baseline and limited the analyses to the 34,501 men

with information on diet and family history. Age-adjusted

characteristics of this group of men, stratified by family his-

tory, are shown in Table 2. Men with and without a family

history were very similar with respect to age, daily dietary

intake, body mass index, and thiazide use.

During 264,835 person-years of follow-up over an 8-yr

period, we documented 795 incident cases of symptomatic

kidney stones. After adjustment for age, a family history of

kidney stones was associated with a increased risk of stone

formation (Table 3). The relative risk for men with a positive

1570 Journal of the American Society of Nephrology

Table 1. Prevalence of personal history of kidney stones,

according to family history of kidney stones at

baseline in 1986 among 37,999 men in the Health

Professionals Follow-up Study

Family His tory of Kid[n (%)]

ney Stones

Yes No

Personal history of kidney

stones (%)

Yes 839 (17.2) 2118 (6.4)

No 4034 (82.8) 31,008 (93.6)

Unadjusted odds ratio = 3.05 (95% CI, 2.79 to 3.32)

Age-adjusted odds ratio = 3. 16 (95% CI, 2.90 to 3.45)

Table 2. Age-adjusted characteristics of the 34,501 men

with no personal history of kidney stones in 1986,

according to family history of kidney stomesa

F amily History of Kidney

VariableStones

(nYes No

395 1 ) (n 30,550)

Age (yr) 54.6 54.5

Daily intake

calcium, dietary (mg) 786 806

calcium, supplemental (mg) 103 105

animal protein (g) 67 67

potassium (mg) 3426 3456

sodium (mg) 3259 3243

magnesium (mg) 355 358

alcohol (g) 11 12

caffeine (mg) 240 239

fluid (ml) 1912 1946

Body mass indexL� 25.5 25.4

Thiazide use (%) 8.8 9.2

a All values except thiazide use are means, standardized

according to the age distribution of the cohort. Nutrient values havebeen adjusted for energy intake.

b The weight in kilograms divided by the square of the height inmeters.

family history of kidney stones, compared with those without,

was 2.64 (95% CI, 2.26 to 3.08). Adjustment for other potential

risk factors only slightly attenuated the increased risk. After

adjusting for age, use of thiazide diuretics, alcohol, body mass

index, and energy-adjusted dietary intake of calcium, animal

protein, sodium, potassium, sucrose, caffeine and fluid, the

relative risk of incident stone formation in men with a positive

family history of kidney stones, compared with those without,

was 2.57 (95% CI, 2.19 to 3.02).

The magnitude of the association between family history and

incident stone formation was significantly greater in men

younger than 60 yr of age (RR = 2.88) than in those men 60

Table 3. Age-adjusted and multivariate relative risks of

symptomatic incident kidney stones among 34,501

men, according to family history of kidney stonesa

Risk

Family History of Kidney Stones

No Yes(n 30,550) (n 3951)

Whole cohort

number of cases 592 203

person-time (yr) 235,177 29,658

age-adjusted RR 1.0 2.64

95% CI referent (2.26 to 3.08)

multivariate RR 1.0 2.57

95% CI referent (2.19 to 3.02)

Age <60 yr

multivariate RR 1.0 2.88

95% CI referent (2.40 to 3.46)

Age �60 yr

multivariate RR 1.0 1.74

95% CI referent (1.18 to 2.57)

a � denotes the relative risk, compared with the group with no

family history of kidney stones, and CI denotes confidence interval.

The multivariate model included age (in 5-yr age categories), useof thiazide diuretics (yes or no), alcohol (seven categories), bodymass index (six categories), supplemental calcium intake (four

categories), and dietary intake of calcium, animal protein, sodiumpotassium, sucrose, caffeine, and fluid (quintile groups).

yr of age and older (RR = 1.74; test for interaction, P = 0.026)

(Table 3).

To examine whether the impact of dietary risk factors varies

by the presence or absence of family history of stones, we

studied the association between nutrient intake and risk of

incident kidney stones, stratified by family history. The results

presented are for the relative risks in the highest quintile of

intake of the specific nutrient, compared with the lowest quin-

tile (Table 4). Family history did not significantly modify the

inverse association between the intake of dietary calcium or

potassium and the risk of incident kidney stones. The risk of

stone formation in men in the highest, compared with the

lowest, quintile of dietary calcium intake was 0.59 (95% CI,

0.35 to 0.98) for men with a family history and 0.72 (95% CI,

0.55 to 0.96) for men without. Supplemental calcium intake

was not associated with risk, and this did not differ by family

history. The risk of stone formation in men in the highest,

compared with the lowest, quintile of potassium intake was

0.40 (95% CI, 0.23 to 0.70) for men with a family history and

0.45 (95% CI, 0.33 to 0.62) for men without. The magnitude of

the risks for animal protein and sodium intake appeared to be

higher in men with a family history, but these were not statis-

tically different from the risks in men without. Caffeine and

fluid intake were significantly inversely associated with risk of

incident stone formation only in men without a family history,

but the estimates were only marginally significantly different

from those with a family history (tests for interaction, �caffejne

= 0.06 and ‘3fluid 0.09).

Family History and Kidney Stones I 571

Table 4. Multivariate relative risks for incident kidney stones in the highest, compared with the lowest, quintile of energy-

adjusted nutrient intake among 34,501 men, according to family history of kidney stones

Nutrient

Daily Intake”Relative Risk” (95% CI) by Family History of

Kidney Stones

HighestQuintile

LowestQuintile

Yes (cases = 203) No (cases = 592)

Calcium, dietary (mg)

Animal protein (g)

Sodium (rug)

Potassium (fig)

Sucrose (g)

Caffeine (mg)

Fluid (ml)

� 1002

�81.4

�3979

�3992

�67

�581

�2544

�555

�53.4

�2369

s2854

s32

�26

�1274

0.59 (0.35 to 0.98) 0.72 (0.55 to 0.96)

1.47 (0.87 to 2.48) 1.15 (0.86 to 1.53)

1.38 (0.83 to 2.29) 0.96 (0.73 to 1.26)

0.40 (0.23 to 0.70) 0.45 (0.33 to 0.62)

1.15 (0.65 to 2.01) 1.06 (0.78 to 1.44)

0.93 (0.57 to 1.51) 0.65 (0.48 to 0.88)

1.00 (0.60 to 1.67) 0.58 (0.42 to 0.79)

Highest

Category

Lowest

Category

Calcium, supplemental (mg) >500 0 0.90 (0.44 to 1.88) 0.94 (0.62 to 1.41)

a For illustrative purposes, cutpoints for the dietary variables are from the 1986 questionnaire. However, the period-specific values were

used for the full 1986-1994 analyses.b The relative risk is the risk of incident kidney stone formation in the men in the highest quintile of intake, compared with the lowest

quintile. The multivariate model included age (in 5-yr age categories), use of thiazide diuretics (yes or no), alcohol (seven categories),body mass index (six categories), supplemental calcium intake (four categories), and dietary intake of the listed variables (quintile groups).

CI, confidence interval.

The results of the logistic models using the raw nutrients

differed very little from energy-adjusted models. The only

substantial change were the relative risks for caffeine-0.72

(95% CI, 0.43 to 1 .20) for men with a family history and 0.73

(95% CI, 0.53 to 0.99) for men without a family history.

24-H Urine Values

The means (± SD) of the 24-h urine values in 369 men

according to family history and case status are shown in Table

5. Among men with incident stones (“cases”), those with a

family history of stones had significantly higher mean 24-h

urinary calcium and citrate excretion. Among men without a

family history, the cases excreted significantly more calcium

and phosphorus, less potassium and total volume, and had

significantly higher calcium oxalate supersaturation. Although

among men with a family history the cases appeared to excrete

more calcium and citrate and to have higher calcium oxalate

supersaturation, the results were not significantly different

from the controls.

DiscussionThese data support previous observations that men who have

suffered from kidney stones are approximately three times

more likely to have a family history of kidney stones. A family

history of kidney stones also substantially increases the risk of

stone formation in men who never have had a stone, indepen-

dent of their dietary intake. The magnitude of the association

between family history and risk of stones was greater in men

younger than 60 yr of age. However, there were no strong

interactions between family history and nutrient intakes. In

particular, family history did not modify the inverse associa-

tion between dietary calcium intake and the risk of incident

stone formation.

Several genetic factors are likely to play a role in calcium

oxalate stone disease and may result in abnormal excretion of

calcium, uric acid. citrate, inhibitors, or promoters (6). The

inheritance of these factors appears to be polygenic (1,6),

although there have been reports of rare families that are

consistent with autosomal monogenic dominant (7) and X-

linked recessive (14) patterns.

Hypercalciuria is the abnormality most commonly identified

with a family history of stones. However, not all hypercalciuric

individuals have a family history of kidney stones and not all

patients with stones and a family history are hypercalciuric.

Calcium restriction has been recommended for some patients

with hypercalciuria and stones, with the goal of decreasing

urinary calcium excretion. However, calcium restriction actu-

ally may increase the risk of stone formation even in patients

with hypercalciuria, potentially because of the increase in

gastrointestinal absorption of oxalate ( I 5). Because our infor-

matiom on 24-h urinary calcium excretion represents only a

subset of our cohort, the interaction between hypercalciuria

and dietary calcium intake cannot be specifically addressed.

Nevertheless, our results suggest that a higher dietary calcium

intake decreases the risk of stone formation to a similar degree

in men with and without a family history. In addition, calcium

supplement use was not associated with risk in either category

of family history. The lack of an association between calcium

supplement use and stone risk may be due to several factors,

such as increased urinary calcium excretion, the timing of

ingestion (typically not with meals), and increased citrate ex-

1572 Journal of the American Society of Nephrology

Table 5. Twenty-four-hour urine values (means ± SD) among 369 men with and without incident kidney stone formation,

according to family history of kidney stonesa

.

Variable

Family History

Yes No

Cases Controls Cases Controls(n 74) (n 13) (n = 209) (n = 73)

Age, 1994 (yr) 57.0 ± 7.2 58.6 ± 7.5 57.5 ± 7.1 60.6 ± 6.8

Body mass index, 1994 (kg/m2) 26.0 ± 2.8 29.1 ± 4.4 26.1 ± 3.6 26.2 ± 3.5

Calcium (mg) 264 ± 1 14” 201 ± 104 229 ± 1 19b,c 184 ± lO2C

Oxalate (mg) 47 ± 14 45 ± 10 46 ± 14 44 ± 12

Sodium (mg) 194 ± 72 211 ± 84 190 ± 70 183 ± 76

Potassium (mg) 82 ± 27 84 ± 26 75 ± 25� 85 ± 26C

Magnesium (mg) 133 ± 44 123 ± 29 127 ± 42 123 ± 36

Uric acid (mg) 741 ± 219 793 ± 165 683 ± 253 709 ± 203

Citrate (mg) 757 ± 330b 684 ± 258 658 ± 291b 640 ± 253

Phosphorus (mg) 1167 ± 354 1176 ± 278 1152 ± 327C ll()4 ± 267C

Volume (L) 1.7 ± 0.7 1.9 ± 0.9 1.6 ± 0.6c 1.8 ± o.8�

Supersaturation

calcium oxalate 2.86 ± 1.35 2.10 ± 1.30 2.56 ± 1.31c 1.82 ± 0.99c

uric acid 3.05 ± 2.05 2.93 ± 2.01 2.88 ± 2.01 2.61 ± 1.72

sodium urate 3.81 ± 2.48 4.29 ± 3.26 3.58 ± 2.39 3.18 ± 2.28

a Twenty-four-hour urine collections were performed after the incident stone event in the cases.b p < 0.05 for comparison of cases with and without family history.C p .< 0.05 for comparison of cases with controls without family history.

cretion as a result of the increased alkaline load of calcium

carbonate.

Hyperuricosuria may also be related to family history. Uric

acid metabolism and excretion may be influenced by inherited

factors ( 16), and men with a gouty diathesis are at increased

risk of stone formation (17). In this study, the mean 24-h uric

acid excretion was higher in both cases and controls in men

with a positive family history, but the differences were not

statistically significant.

Urinary oxalate excretion is influenced by the dietary intake

of oxalate and endogenous oxalate production, but the relative

contribution of these two sources remains controversial. Al-

though there are recognized genetic disorders resulting in a

dramatic overproduction of oxalate, such as primary hyperox-

aluria ( 18), the vast majority of stone formers have urinary

oxalate levels that are modestly elevated or in the normal

range.

Hypocitraturia is also a risk factor for calcium oxalate stone

disease. Surprisingly, urinary citrate levels were significantly

higher in the cases with a family history. It is unlikely that this

difference occurs as a result of treatment because very few men

were being medically treated for their stone disease. We are

unaware of other published data on urinary citrate excretion

and family history.

To decrease the likelihood of stone formation, patients are

routinely advised to increase their urine volume by increasing

their fluid intake. We have previously shown an inverse asso-

ciation between total fluid intake and risk of stone formation

(19). In addition, caffeine use would be expected to result in a

more dilute urine, and thus lower the risk, by interfering with

the action of antidiuretic hormone in the distal nephron (20). It

is unclear why fluid intake and caffeine were not associated

with reduced risk in men with a family history but were

associated in men without a family history. These findings

perhaps provide a clue for how family history might influence

the risk of stone formation.

Despite the strong association between family history and

kidney stones, most individuals who have kidney stones do not

have a family history. In our cohort, 72% of the prevalent cases

and 74% of the incident cases did not have a family history of

kidney stones. Although the magnitude of the risk of incident

stone formation associated with family history was much

higher than for any of the individual dietary factors, the ma-

jority of stone formers did not have a family history. Thus it is

likely that dietary factors-in particular, the intake of calcium,

animal protein, and potassium-play a more important role in

stone formation than family history among most patients who

form stones.

There are potential explanations, other than genetic, for the

observed increased risk associated with a positive family his-

tory. The family history information was collected in 1994,

after all of the cases had occurred. Thus there may have been

reporting bias because individuals who have had a stone may

have been more likely to search out a history in family mem-

bers. However, this is unlikely because our results are consis-

tent with previous reports in which family members themselves

Family History and Kidney Stones 1573

were contacted about their history (I). Another possible expla-

nation is that individuals with a family history were more likely

to share similar environmental exposures that predispose to

stone formation, such as dietary factors (9) or geographic

region (21). However, in our cohort, family history remained

independently associated with risk even after controlling for

dietary factors and geographic region. Biased recall of diet was

avoided because the dietary information was collected before

the diagnosis was made. The potential for information bias is

also unlikely to explain the observed associations between the

dietary factors and risk within strata of family history.

These results are generalizable to men aged 40 yr and older.

The age-specific incidence rates were stable between the ages of

40 to 59 (data not shown), suggesting that these findings may

apply to younger men as well. Whether these results apply to

women remains to be studied, but family history also appears to

be an important risk factor in women (2).

In conclusion, our results support the belief that a family

history of kidney stones substantially increases the risk of stone

formation. Family history does not appear to influence the

association between a high dietary calcium intake and reduced

risk of stone formation. Therefore, dietary calcium restriction

may increase the risk of stone formation, even among imdivid-

uals with a family history of kidney stones.

AcknowledgmentsThis study was supported by research grants DK45362, HL35464,

and CA55075 from the National Institutes of Health.We are indebted to the participants of the Health Professionals

Follow-up Study for their continuing cooperation and to Dr. Sharon

Curhan, Ms. Elaine Coughlan-Havas, Ms. Laura Packer, Mr. Al Wing,Ms. Betsy Frost-Hawes, Ms. Kerry Pillsworth, Ms. Mitzi Wolff, Ms.Jill Arnold, and Mrs. Mira Koyfman for their expert help.

References1. Resnick M, Pridgen DB, Goodman HO: Genetic predisposition

to formation of calcium oxalate renal calculi. N EngI J Med 278:

1313-1318, 1968

2. Trinchieri A, Mandressi A, Luongo P, Coppi F, Pisani E: Famil-

ial aggregation of renal calcium stone disease. J Urol I 39:

478-481, 1988

3. Ljunghall 5, Hedstrand H: Epidemiology of renal stones in a mid-

dle-aged male population. Acta Med Scand 197: 439-445, 19754. Thun Mi, Schober 5: Urolithiasis in Tennessee: An occupational

window into a regional problem. Am J Public Health 81 : 587-

597, 1991

5. Coe FL, Parks JH, Moore ES: Familial idiopathic hypercalciuria.

N EngI J Med 300: 337-400, 1979

6. Goodman HO, Holmes RP, Assimos DG: Genetic factors in

calcium oxalate stone disease. J Urol I 53: 30 1-307, 19957. Baggio B, Gambaro G, Marchini F, Cicerello E, Tenconi R,

Clementi M, Borsatti A: An inheritable anomaly of red-cell

oxalate transport in “primary” calcium nephrolithiasis correct-

able with diuretics. N EngI J Med 314: 599-604, 1986

8. Coe FL, Parks IH: Stone disease in hereditary distal renal tubular

acidosis. Ann Intern Med 93: 60-61, 1980

9. Curhan GC, Willett WC, Rimm EB, Stampfer MJ: A prospective

study of dietary calcium and other nutrients and the risk of

symptomatic kidney stones. N Engl J Med 328: 833-838, 1993

10. Willett WC, Sampson L, Stampfer MI, Rosner B, Bain C, Wits-

chi I, Hennekens CH, Speizer FE: Reproducibility and validity of

a semiquantitative food frequency questionnaire. Am J Epidemiol

122: 51-65, 1985

1 1 . Rimm EB, Giovannucci EL, Stampfer Ml, Colditz GA, Litin LB.

Willett WC: Reproducibility and validity of an expanded self-

administered semiquantitative food frequency questionnaire

among male health professionals. Am J Epidemiol 135: 1 114-

1126, 1992

12. Willett WC: Nutritional Epidemiology, New York, Oxford Uni-

versity Press, I 990

13. D’Agostino RB, Lee MLT, Belanger AJ, Cupples LA, Anderson

K, Kannel WB: Relation of pooled logistic regression to time

dependent Cox regression analysis: The Framingham Heart

Study. Stat Med 9: 1501-15 15, 1990

14. Reinhart SC, Norden AG, Lapsley M, Thakker RV, Pang 1,

Moses AM, Frymoyer PA, Favus MJ, Hoepner JA, Scheinman

SI: Characterization of carrier females and affected males with

X-linked recessive nephrolithiasis. J Am Soc Nephrol 5: 145 1-

1461, 1995

15. Lemann J Jr: Composition of the diet and calcium kidney stones.

N Engl J Med 328: 880-882, 1993

16. Kelley WN, Schumacher HR Jr: Gout. In: Textbook of Rheuma-

tology, 4th ed, edited by Kelley WN, Harris Jr ED, Ruddy 5,

Sledge CB, Philadelphia, W. B. Saunders, 1993, pp 1291-1336

17. Coe FL: Uric acid and calcium oxalate nephrolithiasis. Kidney

mt 24: 392-403, 1983

I 8. Williams HE, Smith LH: Disorders of oxalate metabolism. Am J

Med4S: 715-735, 1968

19. Curhan G, Willett W, Rimm E, Spiegelman D, Stampfer M:

Prospective study of beverage use and the risk of kidney stones.

Am J Epidemiol 143: 240-247, 1996

20. Roberston G: Regulation of vasopressin secretion. In: The Kid-

ney: Physiology and Pathophysiology, 2nd ed, edited by Seldin

D, Giebisch G, New York, Raven Press, 1992

21. Curhan GC, Rimm EB, Willett WC, Stampfer MI: Regional

variation in nephrolithiasis incidence and prevalence among

United States men. J Urol 151: 838-841, 1994