Association between chronic kidney disease and urinary calculus by stone location: a population-based study

Authors


Herng-Ching Lin, School of Health Care Administration, Taipei Medical University, 250 Wu-Hsing St., Taipei 110, Taiwan. e-mail: henry11111@tmu.edu.tw

Abstract

Study Type – Disease prevalence study (cohort design)

Level of Evidence 2a

What's known on the subject? and What does the study add?

Several studies have estimated the potential association of urinary calculus (UC) with chronic kidney disease (CKD). However, previous literature focusing on this issue tended to evaluate the impact of kidney stones alone on incident CKD, with no studies having been conducted investigating the association between CKD and stone formation in other portions of the urological system.

We found that patients with CKD were consistently more likely than comparison subjects to have been previously diagnosed with kidney calculus (odds ratio [OR] 2.10, 95% confidence interval [CI] 1.95–2.27), ureter calculus (OR 1.68, 95% CI 1.51–1.85), bladder calculus (OR 1.49, 95% CI 1.13–1.98), and unspecified calculus (OR 1.89, 95% CI 1.74–2.06). We concluded that there was an association between CKD and UC regardless of stone location.

OBJECTIVE

  • • To explore the association of chronic kidney disease (CKD) with prior kidney calculus, ureter calculus, and bladder calculus using a population-based dataset in Taiwan. Several studies have estimated the potential association of urinary calculus (UC) with CKD. However, previous literature focusing on this issue tended to evaluate the impact of kidney stones alone on incident CKD, with no studies having been conducted investigating the association between CKD and stone formation in other portions of the urological system.

PATIENTS AND METHODS

  • • We identified 21 474 patients who received their first-time diagnosis of CKD between 2001 and 2009.
  • • The 21 474 controls were frequency-matched with cases for sex, age group, and index year.
  • • We used conditional logistic regression analyses to compute the odds ratio (OR) and corresponding 95% confidence interval (CI) as an estimation of association between CKD and having been previously diagnosed with UC.

RESULTS

  • • The results show that compared with controls, the OR of prior UC for cases was 1.91 (95% CI 1.81–2.01, P < 0.001) after adjusting for potential confounders.
  • • Furthermore, cases were consistently more likely than controls to have been previously diagnosed with kidney calculus (OR 2.10, 95% CI 1.95–2.27), ureter calculus (OR 1.68, 95% CI 1.51–1.85), bladder calculus (OR 1.49, 95% CI 1.13–1.98), and unspecified UC (OR 1.89, 95% CI 1.74–2.06).

CONCLUSION

  • • We concluded that there was an association between ckd and UC regardless of stone location.
Abbreviations
CHD

coronary heart disease

CKD

chronic kidney disease

LHID2000

Taiwan Longitudinal Health Insurance Database 2000

NHI

National Health Insurance

OR

odds ratio

PVD

peripheral vascular disease

SLE

systemic lupus erythematosus

UC

urinary calculus

INTRODUCTION

Chronic kidney disease (CKD) is a common and serious disorder, affecting 13% of the adult population [1]. It is associated with a significantly increased risk of hospital admission, morbidity, and death from cardiovascular disease [2]. In addition, CKD can also progress to end-stage renal disease, which results in patients requiring dialysis and/or renal transplantation. Hence, the identification of possible risk factors associated with the development of CKD has become an important focus area for researchers.

Age, obesity, hypertension, diabetes, gout, anaemia, and hyperuricaemia have been established as potential risk factors for CKD [3–7]. Besides these, several recent studies have also considered urinary calculus (UC) to be a significant and independent risk factor for CKD. A population-based historical cohort study by Rule et al. [8] found that when compared with matched control subjects, kidney stone formers had a 51–68% increased risk for CKD as defined by diagnostic codes, and a 25–44% increased risk for CKD as defined by elevated serum creatinine levels. Hippisley-Cox and Coupland [9] recently performed a large population-based historical cohort study and found that women with kidney stones in England and Wales had a 27% increased adjusted risk for moderate to severe CKD.

Besides cohort studies, some other population-based studies have also investigated the association between kidney stones and CKD. Gillen et al. [10] used the Third National Health and Nutrition Examination Survey (NHANES III) of the USA population to compare estimated GFRs between persons with and without a history of kidney stones. They found that a history of kidney stones may reduce kidney function among overweight persons. Moreover, Vupputuri et al. [11] found that patients who had CKD and were identified by diagnostic codes and elevated serum creatinine levels were more likely to report a history of kidney stones on telephone interview when compared with matched community control subjects (odds ratio [OR] 1.9).

Although several studies have estimated the potential association of UC with CKD, previous literature focusing on this issue tended to evaluate the impact of kidney stones on CKD. As far as we know, no study has investigated the association between the stone formation in other portions of the urological system and CKD. Therefore, in order to contribute to the knowledge surrounding this association, this case-control study aimed to explore the association of CKD with prior kidney calculus, ureter calculus, bladder calculus and unspecified urinary calculus using a nationwide population-based dataset.

PATIENTS AND METHODS

We used data retrieved from the Taiwan Longitudinal Health Insurance Database 2000 (LHID2000). Taiwan inaugurated its National Health Insurance (NHI) programme on 1 March 1995. The LHID2000 includes the medical claims data for 1 million beneficiaries, randomly sampled from the year 2000 registry of NHI beneficiaries (n= 23.72 million). The Taiwan National Health Research Institute has validated the representativeness of the LHID2000, and confirmed that it corresponds with the whole population of NHI beneficiaries on sex ad age distribution. The LHID2000 grants researchers the opportunity to trace the medical histories of the 1 million beneficiaries included in this dataset starting from the initiation of the NHI programme. Many researchers have used the LHID2000 to perform and publish studies in internationally peer-reviewed journals.

Because the LHID2000 consists of anonymous secondary data that are routinely released to the public for research purposes, the present study was exempt from full review by the Taipei Medical University Institutional Review Board.

STUDY POPULATION

To select the cases used in this case-control study, we first identified 21 474 patients aged ≥ 18 years who had received their first-time diagnosis of CKD (ICD-9-CM code 585 (chronic renal failure) or 593.9 (unspecified disorder of kidney and ureter)) during ambulatory care visits between January 2001 and December 2009. We assigned the date of their first CKD diagnosis as their index date. We further selected one control for each case from the remaining beneficiaries in the LHID2000. The 21 474 controls were frequency-matched with cases with for sex, age group (18–39, 40–49, 50–59, 60–69, and >69 years), and index year using the SAS program proc surveyselect (SAS System for Windows, Version 8.2). We assured that none of the selected controls had a history of CKD. For controls, we assigned their first use of medical services occurring in the index year as their index date. A total of 42 948 subjects were included in the present study.

EXPOSURE ASSESSMENT

The UC cases used in the present study were identified based on ICD-9-CM codes 592 (calculus of kidney and ureter), 592.0 (calculus of kidney), 592.1 (calculus of ureter), 592.9 (urinary calculus, unspecified), 594 (calculus of lower urinary tract), 594.0 (calculus in diverticulum of bladder), 594.1 (other calculus in bladder), 594.2 (calculus in urethra) 594.8 (other lower urinary tract calculus), and 594.9 (calculus of lower urinary tract, unspecified). To increase diagnostic validity, we only selected patients who had received two or more UC diagnoses before the index date, with at least one being made by a urologist or nephrologist.

STATISTICAL ANALYSIS

The chi-squared test was used to examine the differences between cases and controls in terms of monthly income (New Taiwan Dollars 0–15 840, 15 841–25 000, ≥25 001) geographic location (Northern, Central, Eastern, and Southern Taiwan), and urbanisation level as well as selected medical co-morbidities including hypertension, diabetes, coronary heart disease (CHD), hyperlipidaemia, obesity, gout, anaemia, alcohol abuse/alcohol dependence syndrome, peripheral vascular disease (PVD), and systemic lupus erythematosus (SLE). These medical co-morbidities are all well-documented risk factors for CKD. Furthermore, we used conditional logistic regression analyses (conditioned on sex, age group, and index year) to compute the OR and corresponding 95% CI as an estimation of association between CKD and having been previously diagnosed with UC. We used the conventional P≤ 0.05 to indicate statistical significance in this study.

RESULTS

Of the total 21 474 cases and 21 474 controls, the mean (sd) age was 62.3 (16.4) years, and over one-quarter were aged > 69 years. Table 1 shows the results of the Pearson chi-squared tests used to examine the differences in the distributions of socio-demographic characteristics and medical co-morbidities between cases and controls. After matching for age, sex, and index year, there was no significant difference in monthly income (P= 0.254), urbanisation level (P= 0.314), and geographic region (P= 0.477) between cases and controls. However, cases had a higher prevalence of co-morbidities in the form of hypertension, diabetes, hyperlipidaemia, CHD, PVD, SLE, gout, anaemia, obesity and alcohol abuse/alcohol dependence syndrome than controls.

Table 1. Demographic characteristics and comorbid medical disorders for patients with CKD and comparison group patients at baseline, Taiwan (n= 42 948)
VariablePatients with CKD (n= 21 474)Comparison patients (n= 21 474) P
Total No.Column %Total No.Column %
  1. NT$, New Taiwan Dollar.

Sex:    1.000
 Male11 77454.811 77454.8 
 Female9 70045.2970045.2 
Age, years    1.000
 18–392 1219.921219.9 
 40–492 50911.7250911.7 
 50–593 70317.2370317.2 
 60–694 56721.3456721.3 
 >695 54325.8554325.8 
Monthly income, NT$    0.254
 0–15 8408 20138.2831038.7 
 15 841–25 0009 93646.3994246.3 
 ≥25 0013 33715.5322215.0 
Urbanisation level:    0.314
 15 95627.7607128.3 
 25 96627.8583927.2 
 33 47516.2353916.5 
 43 20714.9324615.1 
 52 87013.4277912.9 
Geographic region:    0.477
 Northern9 43043.9932043.4 
 Central5 30124.7530424.7 
 Southern6 22529.0629229.3 
 Eastern5182.45582.6 
PVD8523.95152.4<0.001
SLE950.4270.1<0.001
Hypertension12 59158.6948744.2<0.001
Diabetes7 51434.9420719.6<0.001
CHD6 35829.6451021.0<0.001
Hyperlipidaemia6 84231.9466721.7<0.001
Obesity1830.81130.5<0.001
Gout4 54121.2239111.1<0.001
Anaemia2 64212.310494.9<0.001
Alcohol abuse/alcohol dependence syndrome890.4570.30.008

Table 2 presents the prevalence of prior UC between cases and controls. In all, 7802 of the 42 948 sampled subjects (18.2%) had been diagnosed with UC before the index date. Prior UC was found among 4940 (23.0%) cases and 2862 (13.3%) controls (P < 0.001). Conditional logistic regression analysis showed that compared with controls, the OR of prior UC for cases was 1.91 (95% CI 1.81–2.01, P < 0.001) after adjusting for monthly income, geographic location, urbanisation level, hypertension, diabetes, CHD, hyperlipidaemia, obesity, gout, anaemia, alcohol abuse/alcohol dependence syndrome, PVD, and SLE. In addition, the relationship between CKD and UC still remained (OR 1.90, 95% CI 1.80–2.00) after excluding patients with bladder calculus.

Table 2. Crude and adjusted ORs for previously diagnosed UC among the sample patients (n= 42 948)
VariableTotal sample (n= 42 948)Comparison patients (n= 21 474)Patients with CKD (n= 21 474)
  1. ***P < 0.001. †OR was calculated by using stratified Cox proportional regression (stratified on age group). ‡Adjustment for patient's monthly income, urbanisation level, geographic region, PVD, SLE, hypertension, diabetes, CHD, hyperlipidaemia, obesity, gout, anaemia, and alcohol abuse/alcohol dependence syndrome.

Presence of UC, n (%)7802 (18.2)2862 (13.3)4940 (23.0)
Crude OR (95% CI)1.001.94 (1.85–2.04)***
Adjusted OR (95% CI)1.001.91 (1.81–2.01)***
Presence of UC (excluding bladder calculus), n (%)7573 (17.6)2767 (12.9)4806 (22.4)
Crude OR (95% CI)1.001.94 (1.85–2.05)***
Adjusted OR (95% CI)1.001.90 (1.80–2.00)***

Table 3 further presents the ORs for prior UC among the sampled subjects by stone location. Stone location was categorised into kidney (ICD-9-CM code 592.0), ureter calculus (ICD-9-CM code 592.1), bladder (ICD-9-CM code 594.0 or 594.1), unspecified (ICD-9-CM code 592, 592.9, 594, 594.2, 594.8, or 584.9), and two or more locations of UC. After adjusting for monthly income, geographic location, urbanisation level, hypertension, diabetes, CHD, hyperlipidaemia, gout, anaemia, obesity, alcohol abuse/alcohol dependence syndrome, PVD, and SLE, cases were consistently more likely to have been previously diagnosed with kidney calculus (OR 2.10, 95% CI 1.95–2.27), ureter calculus (OR 1.68, 95% CI 1.51–1.85), bladder calculus (OR 1.49, 95% CI 1.13–1.98), unspecified calculus (OR 1.53, 95% CI 1.39–1.68), and two or more locations of UC (OR 1.95, 95% CI 1.66–2.28) than controls.

Table 3. Crude and adjusted ORs for previous UC according to stone location among the sample patients (n= 42 948)
VariableTotal sample (n= 42 948)Comparison patients (n= 21 474)Patients with CKD (n= 21 474)
  1. **P < 0.01. ***P < 0.001. †OR was calculated by using stratified Cox proportional regression (stratified on age group). ‡Adjustment for patient's monthly income, urbanisation level, geographic region, PVD, SLE, hypertension, diabetes, CHD, hyperlipidaemia, obesity, gout, anaemia, and alcohol abuse/alcohol dependence syndrome.

Presence of kidney calculus, n (%)3338 (7.8)1135 (5.3)2203 (10.3)
Crude OR (95% CI)1.002.19 (2.03–2.35)***
Adjusted OR (95% CI)1.002.10 (1.95–2.27)***
Presence of ureter calculus, n (%)1675 (3.9)671 (3.1)1004 (4.7)
Crude OR (95% CI)1.001.68 (1.52–1.86)***
Adjusted OR (95% CI)1.001.68 (1.51–1.85)***
Presence of bladder calculus, n (%)209 (0.5)88 (0.4)121 (0.6)
Crude OR (95% CI)1.001.55 (1.17–2.04)**
Adjusted OR (95% CI)1.001.49 (1.13–1.98)**
Presence of ≥2 locations of UC, n (%)711 (1.7)236 (1.1)475 (2.2)
Crude OR (95% CI)1.002.04 (1.74–2.38)***
Adjusted OR (95% CI)1.001.95 (1.66–2.28)***
Presence of UC, unspecified, n (%)1869 (4.4)732 (3.4)1137 (5.3)
Crude OR (95% CI)1.001.58 (1.44–1.74)***
Adjusted OR (95% CI)1.001.53 (1.39–1.68)***

We further analysed the relationship between CKD and treatment method (extracorporeal shockwave lithotripsy, endoscopic intervention, and percutaneous nephrolithotomy) by comparing the treatment methods of both the cases with CKD, and the controls without CKD who all had previous diagnoses of UC (Table 4). We found that of the patients with UC, cases with CKD were more likely to have received endoscopic intervention (OR 2.43, 95% CI 1.68–3.51) and percutaneous nephrolithotomy (OR 1.42, 95% CI 1.06–1.92) than controls without CKD after adjusting for monthly income, geographic location, urbanisation level, hypertension, diabetes, CHD, hyperlipidaemia, gout, anaemia, obesity, alcohol abuse/alcohol dependence syndrome, PVD, and SLE. However, there was no significant difference in the prevalence of having previously undergone extracorporeal shockwave lithotripsy therapy between the cases with CKD and controls without CKD who had prior UC in the present study (OR 0.99, 95% CI 0.88–1.11).

Table 4. Crude and adjusted ORs for previous treatment procedures among the sampled patients with UC
VariableTotal sample (n= 7802)Comparison patients (n= 2862)Patients with CKD (n= 4940)
  1. ***P < 0.001; *P < 0.05. †OR was calculated by using stratified Cox proportional regression (stratified on age group). ‡Adjustment for patient's monthly income, urbanisation level, geographic region, PVD, SLE, hypertension, diabetes, CHD, hyperlipidaemia, obesity, gout, anaemia, and alcohol abuse/alcohol dependence syndrome.

Patients treated with extracorporeal shockwave lithotripsy, n (%)1459 (18.7)542 (18.9)917 (18.6)
Crude OR (95% CI)1.000.98 (0.87–1.10)
Adjusted OR (95% CI)1.000.99 (0.88–1.11)
Patients receiving endoscopic intervention, n (%)282 (3.6)54 (1.9)228 (4.6)
Crude OR (95% CI)1.002.47 (1.71–3.56)***
Adjusted OR (95% CI)1.002.43 (1.68–3.51)***
Patients receiving percutaneous nephrolithotomy, n (%)218 (2.8)62 (2.2)156 (3.2)
Crude OR (95% CI)1.001.47 (1.09–1.98)*
Adjusted OR (95% CI)1.001.42 (1.06–1.92)*

DISCUSSION

After adjusting for socioeconomic factors and medical co-morbidities, we found patients with CKD to be 1.91-times more likely to have previously received a diagnosis of UC than matched controls. We further analysed the association according to stone location, and also found patients with CKD to be consistently more likely to have been previously diagnosed with kidney calculus (OR 2.10), ureter calculus (OR 1.68), bladder calculus (OR 1.49), and unspecified calculus (OR 1.89) than controls.

While several studies have investigated the role of UC in the development of CKD, to date the literature focusing on this issue has tended to ignore the potential contribution of UC in other parts of the collecting system [8–11]. The present study succeeded in detecting novel associations between ureter calculus, bladder calculus, unspecified calculus, and CKD. Furthermore, the present findings are consistent with previous studies reporting that kidney stone formers had increased risks for CKD than unaffected subjects [8–11].

Although the actual mechanisms contributing to the association between UC and CKD are still yet to be elucidated, several risk factors for renal damage in patients with UC have been proposed to include anatomical abnormalities, infection and inflammation with parenchymal scar formation, underlying metabolic disorders, repeated interventions, environmental factors, dietary factors, and molecular or genetic factors [12–14]. For example, urease-producing agents can cause increases in ammonium production and consequently result in alkaline urine, causing tubular lesions, struvite stone formation, urinary stasis, inflammation, and consequent kidney lesions that all contribute to CKD [15–19]. Furthermore, some patients with UC develop deposits in the inner medullary collecting ducts and ducts of Bellini, which produce significant damage to the nephron [20]. Furthermore, stone passage itself causes transient obstruction, and obstruction is a well-established risk factor for renal damage.

The primary strength of the present study lies in its longitudinal database and large sample population, which mitigate the effect of the selection biases inherent in studies using data taken from voluntary registries or hospital-referred study patients. Also, >98% of Taiwan's residents are of Han Chinese ethnicity. While the homogeneity of the population may exempt the present study from potential confounding by race, it also limits the generalizability of our results to other ethnic groups.

The present findings need to be interpreted with caution due to the following limitations. First, the NHI Research Database uses discharge diagnoses provided by treating physicians. As cases are not defined by any standardised criteria, there is the potential for bias due to case misclassification. Secondly, the NHI Research Database lacks clinical information, which prevented us from being able to take either stone or urine composition into consideration in the present analysis. Furthermore, some lifestyle information, e.g. smoking status, alcohol consumption dietary habits, and body mass index, all of which may have contributed to the development of CKD, were also not available through the administrative dataset. Lastly, on account of the present study design we were unable to determine the direction of any causality; and therefore we could not suggest whether UC causes CKD or whether the presence of CKD causes UC. This should be considered in future prospective studies.

Despite these limitations, the present study still provides an association between CKD and UC. Physicians should be aware of this association and appropriate management should be taken to minimise the risk of CKD in patients with UC. We recommend that future studies be conducted in other regions and among other ethnic groups to explore the associations between UC and CKD to reveal general patterns worldwide. Moreover, further research is needed to determine the association between stone composition, stone burden, the treatment associated with UC, and the risk of CKD.

ACKNOWLEDGEMENTS

This study is based in part on data from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health, Taiwan and managed by the National Health Research Institutes. The interpretations and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health, or the National Health Research Institutes.

CONFLICT OF INTEREST

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. There are no competing interests to declare for all authors.