Temporal Trends of Dietary Risk Factors after a Diagnosis of Kidney Stones

Abstract

Background and objectives 

Diet is an important contributor to kidney stone formation, but there are limited data regarding long-term changes in dietary factors after a kidney stone.

Design, setting, participants & measurements 

We analyzed data from three longitudinal cohorts, the Health Professionals Follow-Up Study and Nurses’ Health Study I and II, comparing changes in dietary factors in participants with and without kidney stones during follow-up. The daily intake of dietary calcium, supplemental calcium, animal protein, caffeine, fructose, potassium, sodium, oxalate, phytate, vitamin D, vitamin C, sugar-sweetened beverages, fluids, net endogenous acid production, and Dietary Approaches to Stop Hypertension score were assessed by repeat food frequency questionnaires and computed as absolute differences; a difference-in-differences approach was used to account for temporal changes using data from participants without kidney stones from the same calendar period.

Results 

Included were 184,398 participants with no history of kidney stones, 7095 of whom became confirmed stone formers. Several intakes changed significantly over time in stone formers, with some showing a relative increase up to 8 years later, including caffeine (difference in differences, 8.8 mg/d; 95% confidence interval [95% CI], 3.4 to 14.1), potassium (23.4 mg/d; 95% CI, 4.6 to 42.3), phytate (12.1 mg/d; 95% CI, 2.5 to 21.7), sodium (43.1 mg/d; 95% CI, 19.8 to 66.5), and fluids (47.1 ml/d; 95% CI, 22.7 to 71.5). Other dietary factors showed a significant decrease, such as oxalate (−7.3 mg/d; 95% CI, −11.4 to −3.2), vitamin C (−34.2 mg/d; 95% CI, −48.8 to −19.6), and vitamin D (−18.0 IU/d; 95% CI, −27.9 to −8.0). A significant reduction was observed in sugar-sweetened beverages intake of −0.5 (95% CI, −0.8 to −0.3) and −1.4 (95% CI, −1.8 to −1.0) servings per week and supplemental calcium of −105.1 (95% CI, −135.4 to −74.7) and −69.4 (95% CI, −95.4 to −43.4) mg/d for women from Nurses’ Health Study I and II, respectively. Animal protein, dietary calcium, fructose intake, Dietary Approaches to Stop Hypertension score, and net endogenous acid production did not change significantly over time.

Conclusions 

After the first episode of a kidney stone, mild and inconsistent changes were observed concerning dietary factors associated with kidney stone formation.

Visual Abstract

Introduction

Diet is an important contributor to kidney stone formation. Dietary factors associated with higher risk of kidney stones include higher dietary intake of animal protein, salt, oxalate, fructose, sugar-sweetened beverages, and supplemental vitamin C (^1–6) and lower intake of calcium, fluids, and potassium (^7–11). Although most studies in the past have focused on individual nutrients, recent studies involving dietary patterns and risk of stone formation have been published given the potential interactions and synergism among individual nutrients from food groups that may better represent the overall effects of an individual’s diet (^12–14). A higher adherence to a Dietary Approaches to Stop Hypertension (DASH) dietary pattern (¹²) or to a Mediterranean dietary pattern showed a lower risk of kidney stones (¹⁴), potentially by increasing urinary citrate and other inhibitors, such as magnesium and phytate, as well as urinary volume (¹⁵).

On the basis of current evidence, patients with kidney stones should be advised to increase their fluid intake (¹⁶); ingest an adequate amount of dietary calcium; and limit their intake of animal protein, oxalate, and salt (¹⁷). They also may receive advice to increase dietary potassium in the form of fruits and vegetables (¹⁷). Although the stone recurrence rate is considered high and depends on factors such as previous stone history and type of treatment (¹⁸), dietary changes can potentially reduce the risk of recurrence (²). In addition, nephrolithiasis is a risk factor for cardiovascular disease (¹⁹); hence, adherence to a Mediterranean or a DASH dietary pattern could reduce risk of kidney stones and also contribute to reduction of major cardiovascular events and increased longevity (^20–22).

Despite the evidence that dietary modification can reduce the development of kidney stones, there are limited data evaluating changes in dietary habits after the diagnosis of a kidney stone. Furthermore, whether and in what time frame such dietary factors are modified after a diagnosis of kidney stones are not known. Therefore, we conducted a longitudinal study of three large cohorts, the Nurses’ Health Study I (NHS I), the Nurses’ Health Study II (NHS II), and the Health Professionals Follow-Up Study (HPFS), and analyzed the changes in known dietary risk factors for kidney stones after a kidney stone diagnosis.

Materials and Methods

Study Population

HPFS was established in 1986 with the enrollment of 51,529 health professionals (dentists, optometrists, osteopaths, pharmacists, podiatrists, and veterinarians) who were men between the ages of 40 and 75 years. NHS I was established in 1976 with the enrollment of 121,700 registered nurses who were women between the ages of 30 and 55 years. NHS II was established in 1989 with the enrollment of 116,430 registered nurses who were women between the ages of 25 and 42 years. At enrollment, participants from each cohort completed a questionnaire with detailed information on diet, medical history, and medications. Participants were followed up with biennial questionnaires, providing updated information and newly diagnosed diseases.

In this study, we excluded participants with a history of kidney stones at baseline and participants with a history of malignancy (except for nonmelanoma skin cancer). Participants with missing data for risk factors for kidney stones at baseline were also excluded, whereas those with missing data for just a given time period were skipped until the next time period. A flow diagram of sample selection in each cohort is presented in Supplemental Figure 1.

Assessment of Diet

In 1986 (HPFS and NHS I) and 1991 (NHS II), participants completed a semiquantitative food frequency questionnaire (FFQ) asking about the average annual intake of >130 foods and 22 beverages; subsequently, the information was updated every 4 years. Consumption frequency of each food unit was used to compute intakes of nutrients with data from the US Department of Agriculture, except for oxalate content, which was measured via capillary electrophoresis in the majority of foods from FFQ (²³). All nutrients were energy adjusted to account for total amount of food eaten. FFQs were previously found to be reproducible and valid in these cohorts (^24–27). Updated information on intake of supplements containing calcium, vitamin C, and vitamin D (individually or as a component of a multivitamin) and intake of fluids and caffeine were also derived from FFQs.

Assessment of Other Covariates

Information on age, body mass index, history of diabetes, history of hypertension, and use of thiazides was obtained at baseline and updated every 2 years. Body mass index was calculated as the weight (from a validated questionnaire) in kilograms divided by the square of height in meters (²⁸).

Assessment of Exposure

The primary exposure was a confirmed incident kidney stone event. Participants who reported the occurrence of a kidney stone on a biennial questionnaire received an additional questionnaire about the date of occurrence and associated symptoms (e.g., pain, hematuria). Individuals with a symptomatic stone event were included as kidney stone formers. The self-reported diagnosis was shown to be valid in >95% of self-reported cases in patients who completed the additional questionnaire (¹²).

Assessment of Outcomes

The primary outcomes were the changes in the individual dietary factors from before the diagnosis of the incident kidney stone to after the diagnosis. We computed the absolute changes from before to after (1–4 and 5–8 years) a kidney stone event in the daily intake of the following dietary factors: dietary calcium, supplemental calcium, animal protein, caffeine, fructose, potassium, sodium, oxalate, phytate, vitamin D, vitamin C, sugar-sweetened beverages, and fluids. DASH score and net endogenous acid production (NEAP) were also evaluated.

To obtain estimates that would account for general temporal changes, we calculated the difference in differences (DID) by subtracting absolute changes computed for participants from the same time period who did not form a kidney stone. Participants who developed a kidney stone in the subsequent time period were censored at that time point.

Statistical Analyses

We allocated person-time according to exposure status at the start of each follow-up period. The analysis took into account both time axes, namely time after diagnosis and calendar period. Linear regression models that adjusted for age, body mass index, history of high BP, history of diabetes, and use of thiazides were used to compute DID; clustered SEMs were incorporated to account for repeated measurements on the same participant. Estimates of DID were obtained from pooled cohorts except when an interaction term between the cohort indicator variable and the stone status indicator variable was statistically significant for both time periods, indicating consistent heterogeneity across cohorts; in those cases, estimates were reported separately for each cohort. The SURVEYREG procedure of SAS software version 9.4 (SAS Institute, Cary, NC) was used. A two-tailed P value of 0.05 was considered statistically significant. The research protocol for this study was reviewed and approved by the institutional review board of Brigham and Women’s Hospital, and it has been performed in accordance with the 1964 Declaration of Helsinki.

Results

A total of 184,398 participants were included, with a median follow-up time of 12 years (25th–75th percentiles: 8, 18 years). Baseline characteristics of the study cohorts are shown in Table 1. The number of participants who became stone formers during follow-up was 7095 (1805 from HPFS, 1778 from NHS I, and 3512 from NHS II). Adjusted means for each dietary factor at baseline and in both time periods of follow-up, separated by cohort, are shown in Supplemental Table 1. Results obtained from pooled cohorts are shown in Table 2, and several intakes changed significantly over time. In particular, significant increases regarding caffeine (DID, 8.8 mg/d; 95% confidence interval [95% CI], 3.4 to 14.1), potassium (23.4 mg/d; 95% CI, 4.6 to 42.3), phytate (12.1 mg/d; 95% CI, 2.5 to 21.7), sodium (43.1 mg/d; 95% CI, 19.8 to 66.5), and fluid intake (47.1 ml/d; 95% CI, 22.7 to 71.5) were observed up to 8 years later. Of note, among those with a relative and significant increase over time, such an increase was due to higher intake in stone formers only for potassium and phytate. On the other hand, for sodium, caffeine, and fluid intake, such intakes declined in all groups over time but less in stone formers than in nonstone formers. Thus, there was a significant difference in the magnitude of decrease in sodium, caffeine, and fluid intake between the two groups, such that sodium, caffeine, and fluid intake decreased less in stone formers than in nonstone formers over time. Conversely, some dietary factors showed a relative decrease up to 8 years later, including oxalate (−7.3 mg/d; 95% CI, −11.4 to −3.2) and vitamin C (−34.2 mg/d; 95% CI, −48.8 to −19.6); in the case of vitamin D, the intake showed a significant decrease only up to 4 years later (−18.0 IU/d; 95% CI, −27.9 to −8.0). Dietary calcium, animal protein, fructose intake, DASH score, and NEAP did not change significantly over time. Because of potential heterogeneity by cohort, supplemental calcium and sugar-sweetened beverages servings were reported separately (Table 3). A relative and significant decrease in sugar-sweetened beverages (−0.5; 95% CI, −0.8 to −0.3 and −1.4; 95% CI, −1.8 to −1.0 servings per week for NHS I and NHS II, respectively) and supplemental calcium (−105.1; 95% CI, −135.4 to −74.7 and −69.4; 95% CI, −95.4 to −43.4 mg/d for NHS I and NHS II, respectively) intake was observed in women, whereas the decrease in supplemental calcium intake was significant among men only up to 4 years later (−27.4 mg/d; 95% CI, −42.1 to −12.6). For those dietary factors with a significant decrease over time, the reduction was due to lower intake in stone formers for vitamin C and sugar-sweetened beverages servings; for oxalate, supplemental calcium, and vitamin D, the reduction was due to higher intake in nonstone formers compared with stone formers during follow-up.

Table 1. - Characteristics of the study cohorts at baseline

Baseline Characteristics	Health Professionals Follow-Up Study, n=39,953	Nurses’ Health Study I, n=56,320	Nurses’ Health Study II, n=88,125
Age, yr	54 (10)	52 (7)	36 (5)
Body mass index, kg/m2	25.5 (3.4)	25.2 (4.7)	24.6 (5.3)
Diabetes, %	1188 (3%)	1828 (3%)	819 (0.9%)
Hypertension, %	8030 (20%)	13,629 (24%)	5551 (6%)
Thiazides, %	3476 (9%)	7490 (13%)	1498 (2%)
Dietary calcium, mg/d	800 (305)	720 (253)	886 (305)
Supplemental calcium, mg/d	97 (260)	356 (429)	128 (272)
Median (interquartile range)	0 (0,0)	200 (0,500)	0 (0,162)
Animal protein, g/d	68 (18)	55 (14)	64 (17)
Caffeine, mg/d	244 (253)	283 (227)	244 (223)
Median (interquartile range)	158 (42,377)	238 (97,403)	170 (62,380)
Fructose, g/d	25 (11)	22 (9)	23 (11)
Potassium, mg/d	3444 (718)	3058 (618)	2936 (541)
Sodium, mg/d	3262 (1142)	2848 (1029)	2154 (370)
Oxalate, mg/d	144 (138)	119 (91)	135 (115)
Median (interquartile range)	120 (72,176)	103 (63,148)	109 (62,169)
Phytate, mg/d	941 (390)	708 (269)	783 (241)
Total vitamin C, mg/d	430 (473)	344 (372)	257 (320)
Median (interquartile range)	227 (142,513)	201 (133,368)	158 (106,260)
Total vitamin D, IU/d	407 (311)	342 (252)	390 (263)
Median (interquartile range)	307 (194,546)	266 (157,483)	318 (197,531)
Sugar-sweetened beverages, servings per week	2.4 (4.2)	1.6 (3.6)	3.3 (5.9)
Median (interquartile range)	0.9 (0,3.0)	0.5 (0,0.9)	0.9 (0,3.9)
Total fluid, ml/d	1979 (791)	2010 (716)	2118 (825)
DASH score, points per day	23.7 (5.3)	23.9 (5.3)	23.7 (5.1)
NEAP, mEq/d	31 (15)	27 (12)	36 (15)

Data are reported as means (SD) for continuous variables. DASH, Dietary Approaches to Stop Hypertension; NEAP, net endogenous acid production.

Table 2. - Changes in dietary factors after an incident kidney stone

Dietary Factor	Absolute Change Stone Formers	Absolute Change Nonstone Formers	Difference in Differences (95% Confidence Interval)	P Value
Dietary calcium, mg/d
1–4 yr	7.1	13.7	−6.6 (−15.2 to 2.0)	0.13
5–8 yr	45.4	42.8	2.6 (−7.8 to 13.0)	0.62
Animal protein, g/d
1–4 yr	−2.5	−2.5	−0.1 (−0.6 to 0.4)	0.76
5–8 yr	−5.1	−5.4	0.3 (−0.2 to 0.9)	0.23
Caffeine, mg/d
1–4 yr	−14.7	−13.4	−1.4 (−5.5 to 2.7)	0.50
5–8 yr	−29.1	−37.9	8.8 (3.4 to 14.1)	≤0.001
Fructose, mg/d
1–4 yr	−0.4	−0.4	−0.1 (−0.4 to 0.2)	0.57
5–8 yr	−0.6	−0.4	−0.2 (−0.6 to 0.1)	0.18
Potassium, mg/d
1–4 yr	−0.3	8.3	−8.6 (−24.7 to 7.5)	0.29
5–8 yr	74.2	50.8	23.4 (4.6 to 42.3)	0.02
Sodium, mg/d
1–4 yr	−180.5	−209.8	29.4 (11.0 to 47.8)	0.002
5–8 yr	−124.4	−167.5	43.1 (19.8 to 66.5)	≤0.001
Oxalate, mg/d
1–4 yr	8.8	16.4	−7.5 (−10.8 to −4.3)	≤0.001
5–8 yr	21.0	28.3	−7.3 (−11.4 to −3.2)	≤0.001
Phytate, mg/d
1–4 yr	42.5	35.8	6.7 (−1.3 to 14.6)	0.10
5–8 yr	88.3	76.2	12.1 (2.5 to 21.7)	0.01
Vitamin C, mg/d
1–4 yr	−28.2	0.5	−28.7 (−40.0 to −17.3)	≤0.001
5–8 yr	−24.6	9.6	−34.2 (−48.8 to −19.6)	≤0.001
Vitamin D, IU/d
1–4 yr	37.4	55.4	−18.0 (−27.9 to −8.0)	≤0.001
5–8 yr	107.1	109.7	−2.6 (−15.2 to 9.9)	0.68
Total fluid, ml/d
1–4 yr	−40.9	−98.4	57.5 (36.3 to 78.7)	≤0.001
5–8 yr	−144.0	−191.1	47.1 (22.7 to 71.5)	≤0.001
DASH score, points per day
1–4 yr	−0.2	−0.2	0 (−0.1 to 0.1)	0.90
5–8 yr	−0.4	−0.4	0 (−0.1 to 0.1)	0.99
NEAP, mEq/d
1–4 yr	−2.0	−2.0	0 (−0.4 to 0.4)	0.96
5–8 yr	−5.4	−5.0	−0.4 (−0.9 to 0.1)	0.11

Linear regression models are stratified by time period and adjusted for age, body mass index, history of high BP, history of diabetes, and use of thiazides. Those dietary factors for which the interaction by cohort was significant in both time periods are reported separately by cohort in Table 3. DASH, Dietary Approaches to Stop Hypertension; NEAP, net endogenous acid production.

Table 3. - Changes in dietary factors after an incident kidney stone (cohort specific)

Dietary Factor	Absolute Changes Stone Formers	Absolute Changes Nonstone Formers	Difference in Differences (95% Confidence Interval)	P Value
Supplemental calcium, mg/d
1–4 yr
HPFS	−12.2	15.2	−27.4 (−42.1 to −12.6)	≤0.001
NHS I	−60.9	23.0	−83.9 (−109.9 to −57.9)	≤0.001
NHS II	23.8	90.2	−66.4 (−86.4 to −46.4)	≤0.001
5–8 yr
HPFS	49.5	59.2	−9.7 (−30.5 to 11.1)	0.36
NHS I	−6.9	98.2	−105.1 (−135.4 to −74.7)	≤0.001
NHS II	163.8	233.2	−69.4 (−95.4 to −43.4)	≤0.001
Sugar-sweetened beverages, servings per week
1–4 yr
HPFS	−0.2	−0.2	0 (−0.2 to 0.3)	0.78
NHS I	−1.6	−1.2	−0.4 (−0.7 to −0.2)	≤0.001
NHS II	−3.5	−2.2	−1.2 (−1.6 to −1.0)	≤0.001
5–8 yr
HPFS	−0.5	−0.3	−0.2 (−0.5 to 0.04)	0.10
NHS I	−1.7	−1.2	−0.5 (−0.8 to −0.3)	≤0.001
NHS II	−3.8	−2.4	−1.4 (−1.8 to −1.0)	≤0.001

Linear regression models are stratified by time period and adjusted for age, body mass index, history of high BP, history of diabetes, and use of thiazides. HPFS, Health Professional Follow-Up Study; NHS, Nurses’ Health Study.

Discussion

In this follow-up study, we analyzed long-term changes in dietary factors after a diagnosis of a kidney stone in three large cohorts. We compared participants who developed a kidney stone with those who did not, taking into account general changes over time. We found that certain dietary factors associated with kidney stone formation showed a significant relative change over time after the first diagnosis of a kidney stone, including caffeine, potassium, sodium, oxalate, phytate, supplemental calcium, vitamin C, vitamin D, fluids, and sugar-sweetened beverages servings. We did not observe significant changes regarding dietary calcium, animal protein, fructose intake, DASH score, and NEAP.

The lack of adherence regarding fluid intake was previously described as around 50% among stone formers who received dietary counseling (²⁹). Although we did not observe an increment in fluid intake over time, previous studies have emphasized that even small increases in fluid intake can reduce the risk of new stone formation (¹⁰,³⁰).

Potassium and phytate intake increased significantly more in stone formers but only in the 5- to 8-year period of follow-up. The protective increment of potassium intake (¹,¹¹) was not accompanied by significant changes in DASH score and NEAP over time; both are helpful tools to measure the consumption of fruits and vegetables, consumption of which is currently recommended to reduce risk of stone recurrence (¹²,¹⁴,¹⁷).

Reduction of sodium intake occurred in both groups over time. However, the relative difference was due to lower intake in nonstone formers in the same calendar period. Although there is evidence to recommend a reduction in sodium intake after a stone event (²,¹⁷), we observed a more significant reduction in nonstone formers than stone formers.

Despite the fact that stone formers had a relatively lower oxalate intake years after the stone event, they actually increased their consumption of oxalate but only less so compared with nonstone formers over time. In large observational studies (³), a higher-oxalate diet is associated with a slightly higher risk of an incident kidney stone, and therefore, calcium oxalate stones formers are currently advised to limit their intake of high-oxalate food (¹⁷). However, there are scarce data showing that changing to a low-oxalate diet reduces the risk of stone formation. A retrospective analysis enrolling patients with idiopathic hyperoxaluria demonstrated that dietary management resulted in meaningful reductions of urine oxalate and supersaturation of calcium oxalate during a short-term follow-up (³¹).

Regarding vitamin C intake, we observed a significant reduction over time in stone formers, differing from nonstone formers in the same temporal frame. Also, increases in vitamin D intake were larger in nonstone formers and significant up to 4 years later, both consistent with the current recommendations to not ingest excessive amounts of vitamin C and vitamin D to reduce the risk of stone formation (⁶,¹⁷,³²).

In this study, changes in calcium intake were computed separately as dietary calcium and supplemental calcium. We did not observe significant changes regarding daily dietary calcium intake, which is in accordance with the major recommendation not to follow a low-calcium diet (⁷,⁸,¹⁷). On the other hand, supplemental calcium intake decreased significantly more in stone formers from NHS I and NHS II, which may suggest some concern regarding supplemental calcium intake among women with kidney stones (⁹,³³). In fact, a previous large observational study showed that depending on the total amount of oxalate in the diet and the timing of calcium ingestion, the risk of stone formation is slightly higher among those individuals taking calcium supplements (⁹). Also, in a randomized controlled trial among postmenopausal women, calcium with vitamin D supplementation slightly increased the risk of kidney stones (³⁴).

The beneficial reduction in sugar-sweetened beverages consumption among stone formers from NHS I and NHS II after the diagnosis of a kidney stone is noteworthy (⁵). We observed a marked reduction in sugar-sweetened beverages consumption over time in women (up to −1.4 servings per week compared with nonstone formers) but not in men.

To our knowledge, this is the first large longitudinal cohort study evaluating diet changes in the long-term follow-up of kidney stone formers compared with nonstone formers in the same calendar period. Changes in dietary habits and lifestyle are currently recommended and considered a fundamental tool for the prevention of nephrolithiasis (¹⁷). Prior to this study, there had been a lack of data regarding adherence to changes in dietary factors after the first stone event, even in a short-term follow-up.

We recognize limitations in our study. The generalizability of our results may be limited because these were cohorts of health professionals, and decisions regarding dietary changes could differ from those of less health-savvy individuals. These cohorts are also largely White, and men younger than 40 years were not included. Furthermore, some participants likely had more than one stone, but we do not have that information; therefore, we were not able to examine the possibility that more pronounced changes may have occurred in recurrent stone formers. Also, it would be very useful to demonstrate the effect of the changes in the recurrence rate.

We do not have information regarding stone composition, and future studies correlating dietary changes according to the type of stone would provide very valuable information. Finally, we do not know what advice, if any, participants were given regarding dietary intake.

In conclusion, certain dietary factors associated with kidney stone formation change significantly over time after the first diagnosis of a kidney stone. However, these modest changes are unlikely to substantially reduce the risk of new stone formation. Our findings suggest that stone formers may not be receiving dietary advice or need more help with implementing dietary changes. Given the increasing number of patients with stone disease and the high recurrence rates, more effort should be devoted toward appropriate advice and patient adherence to prevent kidney stone recurrence.

Disclosures

G. Curhan reports employment with OM1, Inc.; consultancy agreements with Decibel Therapeutics; ownership interest in Allena Pharmaceuticals and OM1, Inc.; receiving research funding from Decibel Therapeutics; receiving consulting fees from Allena Pharmaceuticals; serving as an Editor-in-Chief, Emeritus of CJASN; and receiving royalties as a section editor and author for UpToDate. P.M. Ferraro reports consultancy agreements with Allena Pharmaceuticals, Alnylam, AstraZeneca, and BioHealth Italia; receiving honoraria from UpToDate; receiving consultant fees and grant support from Allena Pharmaceuticals, Alnylam, AstraZeneca, BioHealth Italia, and Vifor Fresenius; receiving royalties as an author for UpToDate; serving on the editorial boards of Journal of Nephrology, Kidney and Blood Pressure Research, and Nutrients; and serving as a scientific advisor or member of the board of European Rare Kidney Disease Registry and the board of European Renal Association - European Dialysis and Transplant Association Registry. E. Taylor reports employment with the Veterans Affairs Maine Healthcare System, receiving honoraria from UpToDate (contributor), and receiving research support from the National Institutes of Health. The remaining author has nothing to disclose.

Funding

The work was supported by Foundation for the National Institutes of Health research grants CA167552, CA176726, CA186107, DK094910, DK118057, and DK91417. P.M. Ferraro is a member of European Reference Network for Rare Kidney Diseases project 739532.