|
Comparison of Two Diets for the Prevention of
Recurrent Stones in Idiopathic Hypercalciuria
Loris Borghi, M.D., Tania Schianchi, M.D., Tiziana Meschi,
M.D., Angela Guerra, Ph.D., Franca Allegri, M.D., Umberto Maggiore, M.D., and
Almerico Novarini, M.D.
|
ABSTRACT
Background A
low-calcium diet is recommended to prevent recurrent stones in
patients with idiopathic hypercalciuria, yet long-term data on the
efficacy of a low-calcium diet are lacking. Recently, the efficacy
of a low-calcium diet has been questioned, and greater emphasis has
been placed on reducing the intake of animal protein and salt, but
again, long-term data are unavailable.
Methods We conducted
a five-year randomized trial comparing the effect of two diets in
120 men with recurrent calcium oxalate stones and hypercalciuria.
Sixty men were assigned to a diet containing a normal amount of
calcium (30 mmol per day) but reduced amounts of animal protein (52
g per day) and salt (50 mmol of sodium chloride per day); the other
60 men were assigned to the traditional low-calcium diet, which
contained 10 mmol of calcium per day.
Results At five
years, 12 of the 60 men on the normal-calcium, low-animal-protein,
low-salt diet and 23 of the 60 men on the low-calcium diet had had
relapses. The unadjusted relative risk of a recurrence for the group
on the first diet, as compared with the group on the second diet,
was 0.49 (95 percent confidence interval, 0.24 to 0.98; P=0.04).
During follow-up, urinary calcium levels dropped significantly in
both groups by approximately 170 mg per day (4.2 mmol per day).
However, urinary oxalate excretion increased in the men on the
low-calcium diet (by an average of 5.4 mg per day [60 µmol per day])
but decreased in those on the normal-calcium, low-animal-protein,
low-salt diet (by an average of 7.2 mg per day [80 µmol per day]).
Conclusions In men with
recurrent calcium oxalate stones and hypercalciuria, restricted
intake of animal protein and salt, combined with a normal calcium
intake, provides greater protection than the traditional low-calcium
diet.
Idiopathic hypercalciuria is an important1 and
common2
risk factor for the formation of stones, and uncontrolled
hypercalciuria is a cause of recurrences.3
Thiazides can reduce urinary calcium excretion,4 but
since calcium excretion depends in part on diet,5
initial attempts to decrease hypercalciuria should involve dietary modification.
Since most patients with hypercalciuria have intestinal hyperabsorption
of calcium,6
it is common clinical practice to recommend a low-calcium diet.
However, there are no long-term data on the efficacy of this
approach.
Short-term studies have shown that a low calcium intake
significantly reduces urinary calcium excretion but can cause a
deficiency of calcium and an increase in urinary oxalate.7,8
Curhan et al.9
reported that among men without a history of nephrolithiasis, those
with a high intake of calcium (>26.2 mmol per day) had a 34
percent lower risk of stone formation than did those with a low
calcium intake (<15.1 mmol per day), a finding that makes the
protective efficacy of a low-calcium diet doubtful.10
This observation was later confirmed in women.11
Moreover, studies have shown that animal protein12,13,14,15,16
and salt17,18,19,20,21,22
also have a considerable influence on calcium excretion.
We compared the efficacy of the traditional low-calcium diet with
that of a diet containing a normal amount of calcium but reduced
amounts of animal protein and salt. Increased consumption of water
was recommended with both regimens.
Methods
Study Population
Men referred to our outpatient department were eligible for the
study if they met all the following criteria: idiopathic hypercalciuria
(urinary calcium excretion, >300 mg per day [7.5 mmol per day])
on an unrestricted diet, recurrent formation of calcium oxalate
stones (at least two documented events — that is, colic episodes
with expulsion of stones or radiographic evidence of retained
stones), no known condition that is commonly associated with calcium
nephrolithiasis (e.g., primary hyperparathyroidism, primary
hyperoxaluria, enteric hyperoxaluria, bowel resection, inflammatory
bowel disease, renal tubular acidosis, sarcoidosis, or sponge
kidney), no previous visit to a stone disease center, no current
treatment for the prevention of recurrent stones except for the
advice to increase water intake, and residence in the area of Parma,
Italy.
Eligibility was determined after a run-in period of two to three
months,23,24
during which the cause of stone formation was determined. Each
patient was seen at least three times during the run-in period.
Ultrasound and radiologic studies and serum measurements were
performed, as well as urinalysis, culture, and chemical measurements
in two 24-hour urine specimens, while the men remained on an
unrestricted diet. All patients were clinically evaluated by one of
us.
Eligible men were asked whether they were willing to comply with
the assigned dietary regimen for at least five years. They received
detailed information about the risk factors for urinary stones, with
a focus on the role of calcium, animal protein, and salt in the
diet. They were informed that the purpose of the trial was to
determine which of the two diets under study was more effective.
Randomization
After the run-in period, the men who had given written informed
consent were randomly assigned to one diet or the other. The treating
physicians assigned the men to the dietary regimens on the basis of
a random-number sequence (an odd number for the low-calcium diet and
an even number for the diet containing a normal amount of calcium
and reduced amounts of animal protein and salt). The sequence was
generated by one of us, who enclosed the numbers indicating the
assignments in sealed, numbered envelopes.
Dietary Regimens
The men assigned to the low-calcium diet were instructed to avoid
milk, yogurt, and cheese so that calcium intake would be reduced to
approximately 10 mmol per day. As part of our routine clinical
practice, we also advised the men to avoid consuming large amounts
of oxalate-rich foods (e.g., walnuts, spinach, rhubarb, parsley, and
chocolate).
The other dietary regimen was more complex and specific (Table 1). The
men assigned to this regimen were given written explanations and
detailed information designed to help them comply with the regimen.
As compared with the typical diet in our region,25
this diet was low in protein, particularly that of animal origin, and
low in salt, with a normal-to-high intake of calcium. We also
advised the men on this diet to avoid consuming large amounts of
foods that are rich in oxalate. Patients who found the diet to be
too low in calories were instructed to increase their consumption of
bread, pasta, vegetables, and fruit rather than their consumption of
meat or fish. Both diets included 2 liters of water per day in cold
weather and 3 liters per day in warm weather. Moderate consumption
of wine, beer, carbonated beverages, and coffee was allowed. Further
information on the dietary instructions is available as Supplementary
Appendix 1 with the full text of this article at
http://www.nejm.org.
|
|
Data Collection and Follow-up
Twenty-four-hour urine specimens were obtained at base line (with
values documented as the average of the two sets of measurements performed
before randomization), one week after randomization, and at yearly
intervals during the five years of the study. Urinary volume was
measured as a marker of liquid consumption, sodium (measured by
atomic-absorption spectrophotometry) as a marker of salt intake,
urea (measured by the urease method) as a marker of total protein
intake, and sulfate (measured by ion chromatography) as a marker of
animal-protein intake. Calcium excretion was measured by
atomic-absorption spectrophotometry, oxalate excretion by ion
chromatography, and creatinine excretion by the Jaffe reaction. The
urine specimen obtained one week after randomization was analyzed to
check compliance with the dietary regimen. The ratio of creatinine
excretion to body weight was used to verify that the urine had been
collected correctly.
The relative calcium oxalate saturation was measured with the
use of the Equil computer program at base line and after the first
week of the diet. Subsequently, the relative calcium oxalate saturation
was estimated according to a formula obtained by regression analysis
with the use of data from previous studies.23,24
Outcome Measures
The primary outcome measure was the time to the first recurrence
of a symptomatic renal stone or the presence of a radiographically identified
stone (see below). In the event of a recurrence, the treatment was
considered to have failed, and the patient was withdrawn from the
trial. If there were no recurrences, patients were followed until
the fifth annual visit (month 60). Patients who required treatment
with thiazides or allopurinol for conditions such as hypertension or
gout were withdrawn from the trial.
Recurrences were considered to be either silent or symptomatic.
Silent recurrences were diagnosed on the basis of renal ultrasound and
abdominal flat-plate examinations performed at yearly intervals. If
renal stones were detected, stratigraphy (thin-plane radiography) was
also performed. The imaging studies were performed by a central
radiologic service, and the radiologist had no knowledge of the
trial or the group assignments. A recurrence was classified as
silent if a previously unreported stone was detected in the absence
of symptoms. A symptomatic recurrence was defined as typical renal
colic, an episode of hematuria, or the expulsion or removal of a
previously undiscovered stone. If a symptomatic recurrence was
documented on the basis of renal colic or hematuria, the recurrence
had to be confirmed radiographically.
Secondary outcome measures included changes in calcium and oxalate
excretion, the calcium oxalate product, and the relative calcium oxalate
saturation.
Statistical Analysis
The analysis was based on the intention-to-treat principle. We
used Kaplan–Meier analyses to determine the cumulative incidence of
recurrent stones, and we used Cox proportional-hazards regression to
determine the crude and adjusted relative risks of recurrence. Analyses
were performed with Stata software (version 7, Stata, College
Station, Tex.). Before the study, we estimated that an overall
sample of 120 men was required for 80 percent power at a
significance level of 0.05 to detect a difference of 25 to 50
percent in the risk of a recurrence between the two study groups,
using a two-sided log-rank test.
Although we had not previously planned to do so, we adjusted the
relative risk of a recurrence for clinical characteristics known to
be strong predictors of the likelihood of a recurrence26
— namely, the total number of stones formed previously and the
number of episodes of renal colic in the previous year. In addition,
we tried to determine whether the effects of dietary treatments
varied according to the severity of the disease. To this end, we
established a subgroup of men at highest risk — those in the highest
decile for either of the two predictors of a recurrence. The men at
highest risk (23 of 120, or 19.2 percent) were those with a history
of five or more episodes of colic in the year before randomization,
10 or more stones formed (as documented on the basis of expulsion or
radiography) before randomization, or both. The highest-risk men
tended to have higher base-line urinary indexes, such as higher
levels of oxalate, calcium, sulfate, sodium, and urea, than the
other men, though they also had higher urinary volume. We then
performed an analysis with a Cox model that included an interaction
term for dietary group and the highest-risk category.
For the analysis of the urinary indexes, we compared the two groups
with respect to the absolute change from the base-line value at each
time point. These comparisons were carried out with use of the
Mann–Whitney test. Base-line continuous variables were compared with
use of the Mann–Whitney test and Student's t-test whenever
appropriate; categorical variables were compared with use of
Fisher's exact test.
All data are expressed as means ±SD. A P value of less than
0.05 was considered to indicate statistical significance. All
reported P values are two-sided.
Results
A total of 120 men were enrolled in the study between June 1993
and December 1994, and 60 men were assigned to each diet. Seventeen men
did not complete the study (Figure 1). Of
these 17, 3 assigned to the normal-calcium, low-protein, low-salt
diet withdrew because they did not want to continue with the diet; 7
assigned to the low-calcium diet withdrew because of hypertension, a
possible adverse effect of low calcium intake.27
The base-line demographic and clinical characteristics of the two
groups were similar (Table 2).
|
|
Twenty-three of the 60 men on the low-calcium diet and 12 of the 60
on the normal-calcium, low-protein, low-salt diet had recurrences.
The cumulative incidence of recurrent stones in the two groups is
shown in Figure 2.
The relative risk of a recurrence among the men in the
normal-calcium, low-protein, low-salt group, as compared with the
men in the low-calcium group, was 0.49 (95 percent confidence
interval, 0.24 to 0.98; P=0.04). After adjustment for the total
number of stones formed before randomization and the number of colic
episodes in the year before randomization, the relative risk of a
recurrence was 0.37 (95 percent confidence interval, 0.18 to 0.78;
P=0.006). Further adjustment for the remaining base-line
characteristics did not change the estimate of the relative risk
(data not shown). The incidence of recurrent stones differed
significantly between the two groups only late in the follow-up
period (Figure 2).
As the stratified analysis in Figure 3 shows,
this delayed effect was due to early recurrences in the highest-risk
patients, regardless of the diet to which they were assigned.
|
|
Table 3 shows
the values for the urinary variables throughout the follow-up
period. The 24-hour urinary volume increased to a similar extent in
the two groups. As expected, urinary excretion of sodium, urea
nitrogen, and sulfate did not change with the low-calcium diet,
whereas all three indexes decreased with the normal-calcium,
low-protein, low-salt diet. The decrease in these indexes reflects
dietary compliance, which was excellent in the first week and still
fairly good, although somewhat reduced, during follow-up.
|
As shown in Table
3, calcium excretion decreased with both diets (by approximately
170 mg per day [4.2 mmol per day]). The calcium oxalate product and
the relative calcium oxalate saturation decreased with both diets,
although the reduction was greater with the normal-calcium,
low-protein, low-salt diet. The main difference between the two
diets was oxalate excretion, which increased with the low-calcium
diet (by approximately 5.4 mg per day [60 µmol per day]) but
decreased with the normal-calcium, low-protein, low-salt diet (by
approximately 7.2 mg per day [80 µmol per day]). There were no
differences in dietary compliance between the men who had recurrent
stones and those who did not, irrespective of the diet.
Discussion
This study shows that a diet with a normal amount of calcium but
with reduced amounts of animal protein and salt is more effective
than the traditional low-calcium diet in reducing the risk of
recurrent stones in men with idiopathic hypercalciuria. The
difference appears to be due to the different effects of the two
diets on oxalate excretion.
Studies extending short-term investigations8 have
shown that a low-calcium diet has long-term efficacy in reducing
calcium excretion. However, this diet may cause an increase in
urinary oxalate excretion through increased intestinal absorption, resulting
from the low level of calcium available to form a complex with oxalate
in the intestinal lumen.28,29 In
terms of saturation, the increase in oxalate tends to be offset by
the reduction in calcium, but the concurrent increase in urinary
volume causes a substantial reduction of the calcium oxalate molar
product and, hence, of the relative calcium oxalate saturation.
We found that a diet with a normal amount of calcium but reduced
amounts of animal protein and salt resulted in a reduction in calcium
excretion that was, on the whole, equivalent to that associated with
a low-calcium diet. Indeed, after the first week of treatment, the
drop in urinary calcium excretion was more marked with this diet
than with the low-calcium diet. Subsequently, this difference tended
to disappear, probably because of a partial reduction in compliance.
The decrease in urinary calcium excretion, despite normal calcium
intake, is probably the consequence of the combined tubular action
of the decreased intake of salt and animal protein, a phenomenon
previously documented in short-term studies.12,13,14,15,16,17,18,19,20,21,22
The other important result of the normal-calcium, low-protein,
low-salt diet was the consistent reduction in urinary oxalate excretion.
The explanation for the reduction in oxalate excretion with the
normal-calcium diet is the converse of the explanation for its
increase with the low-calcium diet. With the normal-calcium diet,
more calcium is available in the intestinal lumen to form a complex
with oxalate, thus reducing its absorption — a phenomenon reported
in short-term studies.30,31 In
addition, the reduced intake of protein may lower the endogenous
synthesis of oxalate.32
The normal-calcium, low-protein, low-salt diet decreases urinary
excretion of both calcium and oxalate, which in combination with
an increase in urinary volume causes a marked reduction in the
calcium oxalate molar product and in the relative calcium oxalate
saturation. These effects may explain the 50 percent reduction in
the risk of a recurrence among the men assigned to this diet, as
compared with those assigned to the low-calcium diet. However, this
advantage was evident only after several years of follow-up. We
speculate that the early advantage of the normal-calcium,
low-protein, low-salt diet over the low-calcium diet might have been
obscured by the enrollment of men who were at high risk for an early
recurrence. This interpretation is consistent with that of Parks and
Coe,26
who speculated that at the start of treatment, patients at high risk
may have stones that are too small to be seen on radiographs but
that grow and are later identified as new stones.
Because the patients enrolled in our trial came to us with an
explicit request to receive dietary treatment for the prevention of
recurrences, it would not have been possible or ethical to include a
control group. However, several other studies have included a
placebo group or a conservative-treatment group,33,34,35,36,37,38,39,40,41
and in all these studies, the risk of a recurrence among patients receiving
placebo or conservative treatment was higher than the risk with
either diet in our study. Moreover, in studies of the natural
history of the disease,42,43,44
the risk of a recurrence at five years was approximately 50 percent,
which is higher than the risk with either diet in our trial.
While our study was in progress, the results of a trial that examined
the protective effect of a diet characterized by low levels of
animal protein and high levels of fiber were reported.45
The authors concluded that this regimen was not more beneficial than
the simple advice to increase the intake of liquids. However, this
study differed from ours in several ways. The subjects were patients
with a first episode of nephrolithiasis, only 17 percent of whom had
hypercalciuria. The dietary prescription did not include restricted
salt intake, and there was little control of calcium intake.
Moreover, compliance with the diet was poor.
In conclusion, our study suggests that a diet characterized by
normal calcium, low animal protein, and low salt levels is more
effective than the traditional low-calcium diet for the prevention
of recurrent stones in men with idiopathic hypercalciuria. We
speculate that this type of diet will be of greatest value when it
is started early in the course of the disease.
Supported in
part by grants from the University of Parma and the Italian Ministry
for Universities and for Scientific and Technological Research.
We are indebted to Dr. Maurizio Rossi of the Department of
Pedagogic Sciences at the University of Parma for his valuable
assistance with the computerized data base.
Source Information
From the Departments of Clinical Sciences (L.B., T.S., T.M., A.G.,
F.A., A.N.) and Internal Medicine and Nephrology (U.M.), University of Parma,
Parma, Italy.
Address reprint requests to Dr. Borghi at the Department of
Clinical Sciences, University of Parma, Via Gramsci 14, 43100 Parma, Italy, or
at [log in to unmask].
References
This article has been cited by other
articles:
Edward E.
Rylander, M.D.
Diplomat American
Board of Family Practice.
Diplomat American
Board of Palliative Medicine.