Pediatr Nephrol (2011) 26:21732178 DOI 10.

1007/s00467-011-1927-5

ORIGINAL ARTICLE

Body fat composition and occurrence of kidney stones in hypercalciuric children

Rose Ayoob & Wei Wang & Andrew Schwaderer

Received: 26 March 2011 / Revised: 12 May 2011 / Accepted: 17 May 2011 / Published online: 10 June 2011 # IPNA 2011

Abstract In the last 10 years, the incidence of kidney stones has increased in the pediatric population, and this rise has been paralleled by a significant increase in pediatric obesity rates in the USA. The purpose of this study was to evaluate percentage body fat (%BF) measured by dual energy X-ray absorptiometry (DXA) in hypercalciuric children with and without kidney stones. A retrospective chart review was performed on children with idiopathic hypercalciuria based on a 24-h urine calcium excretion of >4 mg/kg/day or >200 mg/day who had undergone DXA scanning. Patients were then classified by sex and by %BF (3 categories; normal: <27% girls, <21% boys; at risk for obesity: 2736% girls, 2130% boys; obese: >36% girls, >30% boys). The 20032004 NHANES data were used as a control. Fifty patients (24 males) were analyzed, of whom 26% were assessed as having a normal %BF, 44% as being at risk for obesity, and 30% as being obese. Children with an increased %BF had a significantly higher occurrence of kidney stones (p =0.03) than those with a normal %BF. No significant differences were noted in 24-h urine chemistries between the groups. In conclusion, an increased %BF was associated with an increased occurrence of kidney stones in children with idiopathic hypercalciuria.

Keywords Pediatrics . Obesity . Urolithiasis

Introduction The prevalence of symptomatic kidney stones among the general population is between 5 and 10%, with a slight male predominance [1, 2]. In the USA, it is estimated that kidney stones account for 1 per 1,000 to 1 per 7,600 pediatric hospital admissions each year [3]. Hypercalciuria is the predisposing metabolic abnormality most frequently identified in the evaluation of kidney stones [1, 2]. Over the last decade an increasing incidence of kidney stones in children and adolescents has been reported [4]. Concurrent with these increasing rates of kidney stones, the rates of pediatric obesity have increased dramatically. In the USA, the obesity rate has increased threefold in the last 25 years [5]. The association between obesity and kidney stones has been well documented in the adult literature [6, 7]. While body mass index (BMI) has been a useful screening measure for obesity, it is not considered to be the most accurate representation of true body fat. Studies have demonstrated that percentage body fat (%BF) measured by dual energy X-ray absorptiometry (DXA) is a more accurate measure of total and regional body composition [8]. To date, no studies in adult or pediatric populations have examined the association between body composition, reported as DXAmeasured %BF, and the occurrence of kidney stones.

R. Ayoob : A. Schwaderer (*) Pediatric Nephrology, Nationwide Childrens Hospital, Columbus, OH, USA e-mail: Schwaderer.5@osu.edu R. Ayoob e-mail: Rose.Ayoob@nationwidechildrens.org W. Wang Biostatistics Core, Nationwide Childrens Hospital Research Institute, Columbus, OH, USA

Methods Study population Following approval from the Institutional Review Board at Nationwide Children's Hospital (NCH) (IRB06-00475), we performed a retrospective

2174

Pediatr Nephrol (2011) 26:21732178

chart review of patients seen in the NCH Pediatric Nephrology Clinic from January 1, 2000 to December 31, 2008. The patients were initially evaluated by one of two physicians using a consistent protocol. Inclusion criteria were patients aged 817 years with a diagnosis of idiopathic hypercalciuria (IHC) and results from a DXA scan performed at NCH. The patients diagnosis of IHC was based on a 24-h urine collection with a urine calcium excretion of >4 mg/kg/day or >200 mg/day [9, 10]. For patients with multiple urine studies and DXA scans, the earliest urine collection with a corresponding DXA scan was used in the data analysis. Children with chronic immobilization, anatomical defects of the lumbar spine, hypercalcemia, history of malignancy, and concurrent conditions, medications, or diets that could independently result in secondary hypercalciuria or increased %BF were excluded. Control population Data from the 20032004 National Health and Nutrition Examination Survey (NHANES) for children aged 817 years were used as the control population (http://www.cdc.gov/nchs/nhanes.htm). The Caucasian NHANES data set was used to match the racial category of the NCH hypercalciuria patients. DXA scan The DXA scans were obtained at our institution to measure bone mineral density (BMD) in children with kidney stones and hypercalciuria based on their association with low BMD noted in previous publications [69]. BMD and %BF measurements were obtained using the institutions DXA scanner (Hologic-4000 densitometer with Hologic Discovery software ver. 11.2; Hologic, Waltham, MA). Urine chemistry A stone risk panel (Litholink Corp, Chicago, IL) was performed on all 24-h urine samples collected from study patients. The urine parameters analyzed included calcium, citrate, uric acid, oxalate, magnesium, phosphorus, pH, and volume. Urine chemistries were adjusted for body size when appropriate. Data analysis The %BF measurements reported in the NHANES group were obtained using DXA Hologic Discovery software ver. 12.1 [11]. Previously published studies by Shypailo [12] describe the various effects of different DXA software versions on %BF in children. For consistency, the %BF values recorded for our NCH patients were corrected for the different software version [12]. Patients were then classified by gender and %BF as normal (<27% girls, <21% boys), at risk for obesity (2736% girls, 2130% boys), or obese (>36% girls, >30% boys) according to the %BF limits suggested for pediatric populations [12, 13]. Children categorized as at risk for obesity and obese by corrected %BF were considered to have an

increased %BF for gender. Each patients height and weight, measured at the time of the DXA scan, were used to calculate BMI (kg/m2) values. Age- and gender-specific BMI reference values developed by the Centers for Disease Control and Prevention (CDC) were used to calculate BMI and classify children as obese (BMI 95th percentile), overweight (BMI 85 94th percentile), or normal (BMI 84th percentile) [14, 15]. Children classified as overweight or obese were considered to have an increased BMI for age and gender. Statistical analysis The MIANALYZE procedure in the Statistical Software Analysis (SAS) ver. 9.1.3 (SAS Institute, Cary, NC) was used for analysis. A regression analysis was performed to compare the %BF values of the NCH and the NHANES samples controlled by age and gender for Caucasian children. The remaining comparisons were evaluated with the Students t-test with statistical significance set at a p < 0.05.

Results NCH patient demographics and body composition results Of those children screened in the NCH Pediatric Nephrology clinic, 50 (24 males) with a mean age of 11.6 years (range 8.216.5 years) met the inclusion criteria. Approximately 25% of the patients with results from a 24-h urine collection and who met the study criteria for idiopathic hypercalciuria did not have a complete set of DXA measurements. Reasons for the incomplete DXA measurements included parental refusal, failure of patients to follow up with scheduled DXA scans, and poor follow-up in our clinic. Such patients did not differ in age, sex, or clinical history from those with completed DXA scans. No significant differences were found in age, BMI or BMI z-score mean values between males and females (Table 1). However, females had a significantly higher %BF than males (32.8 vs. 25.2%; p =0.008; Table 1). Patients were classified into obesity categories based on gender and %BF. Thirteen children (10 males) were placed in the normal group,, 22 (6 males) were categorized as at risk for obesity, and 15 children (8 males) were in the obese group. Approximately 74% (37/50) of the patients had an increased %BF. The BMIs of the NCH patients with kidney stones were compared with those of children with ICH alone. No significant difference was found between children with kidney stones and those with IHC alone when categorized by BMI percentiles as normal or increased (p =0.62; Fig. 1).

Pediatr Nephrol (2011) 26:21732178 Table 1 Body composition and patient demographics for the Nationwide Childrens Hospital hypercalciuric patients
# Patients
25 20
Normal BMI Increased BMI

2175

Body composition and patient demographics Total number of patients Age (years) Percentage body fata BMI (kg/m2) BMI (z-score) *p <0.01 vs. males

Overall

Male

Female

15 10 5

50 11.62.5 2910.3 204.3 1.0281.85

24 11.82.6 25.211.2 20.95 1.22

26 11.42.5 32.88 * 19.23.6 11.7

0 Stones IHC

Fig. 2 Occurrence of kidney stones in the Nationwide Childrens Hospital (NCH) hypercalciuric pediatric patients according to body mass index (BMI) category

Data are presented as mean valuesstandard deviation (SD). BMI, Body mass index
a

Percentage body fat (%BF) is reported for Nationwide Childrens Hospital (NCH) patients corrected for dual energy X-ray absorptiometry (DXA) software version

(n =21) and all NCH males with hypercalciuria alone (n =13) had a significantly lower mean %BF than the NHANES patients when controlling for age and gender (Table 3).

Association of body composition and urine chemistries with kidney stone occurrence In the NCH study population, 58% (29/50) had radiographic evidence of kidney stones and 42% (21/50) had hypercalciuria alone. Of those with a history of stones, 62% (18/29) were female. The mean age of children with hypercalciuria alone was not significantly different from those with kidney stones (11.2 vs. 11.8 years, respectively). When compared to children with a normal %BF, children with an increased %BF had significantly more kidney stones (p =0.03; Fig. 2). No significant differences were found in the 24-h urine chemistries between obesity categories (Table 2). NCH study population versus 20032004 NHANES control population The comparison of mean %BF results for the NCH study group and NHANES group revealed no significant differences between the two groups overall or between genders (Table 3). All NCH patients with hypercalciuria alone

Discussion Compared to the adult population, a higher proportion of pediatric patients have an identifiable risk factor favoring stone formation. As with any disease process, understanding the pathophysiology and risk factors for stone formation and recurrence is crucial in patient management. Primary or idiopathic hypercalciuria is considered the most common cause of calcium-containing stones [1, 2]. In the adult literature, the relationship between obesity and the development of kidney stones has been extensively researched. An association between increasing BMI and increasing urinary excretion of sodium, calcium, uric acid, and citrate has been reported [16]. Also, a higher prevalence of uric acid stones has been reported in obese adult stone formers than in leaner subjects [6, 17, 18]. Few published studies, however, have examined body composition and risk factors for kidney stone formation among the pediatric population. Multiple risk factors may result in kidney stones, including elevated uric acid and/or oxalates. Our study focused on hypercalciuria since (1) hypercalciuria is the most common metabolic abnormality identified in the evaluation of urolithiasis in children [19]; (2) for the initial evaluation of %BF and nephrolithiasis, a narrow phenotype was desired; (3) DXAs and subsequent %BF measurements were routinely obtained to monitor for low bone density in hypercalciuria but not for other urine risk factors. In the USA, calcium phosphate and calcium oxalate stones are the most common types of stones found in children, accounting for >75% of cases [19]. Hyperuricosuria may be found in children with stones due to metabolic disorders, such as LeschNyhan syndrome, type I glycogen storage disease, or myeloproliferative disorders. Gout is an uncommon cause of kidney stones in children. Uric acid stones may be

25

Normal %BF Increased %BF

# Patients

20 15 10 5 0 Stones IHC

Fig. 1 Occurrence of kidney stones in the NCH hypercalciuric pediatric patients according to obesity category. *p <0.05 vs. normal percentage body fat (%BF). IHC Idiopathic hypercalciuria

2176

Pediatr Nephrol (2011) 26:21732178

Table 2 Urine chemistries for Nationwide Childrens Hospital hypercalciuric pediatric patients by obesity category %BF categories SS CaOx SS CaP pH Ca (24/kg) 6.12.1 5.31.3 5.21.5 0.26 Ca/Cr (24 h) 285255 26061 28681 0.18 Cit/Cr (24 h) 535327 613282 653331 0.60 Ox UA P (24/kg) 239 217 207 0.54 Mg (24/kg) 3.10.6 2.60.8 2.50.8 0.08 Volume (ml) 0.980.45 0.960.83 0.970.53 0.89

Normal At Risk Obese P value

11.4 4.5 11.44.9 137 0.66

2.82 3.61.5 3.21.7 0.35

6.40.6 6.50.4 6.20.5 0.26

5023 4320 3612 0.23

0.570.17 0.640.22 0.570.25 0.50

Data are presented as mean valuesSD SS CaOx, Supersaturation of calcium oxalate; SS CaP, supersaturation of calcium phosphate; pH, 24-h urine pH; Ca, urine calcium per kilogram; Ca/Cr, calcium/creatinine (mg/day); Cit/Cr, urine citrate/creatinine (mg/day); Ox, urine oxalate; UA, urine uric acid; P, urine phosphate per kilogram (g/kg/day); Mg, urine magnesium per kilogram (mg/kg/day)

seen in 510% of children placed on a ketogenic diet for seizure control. Additional factors that promote uric acid precipitation include supersaturation of the urine with uric acid, low urinary pH, low fluid intake, uricosuric drugs, and low urinary citrate. In the adult literature, an increasing BMI has been associated with decreased urine volume and pH as well as increased urinary oxalate. This relationship has not been clearly established in the pediatric population. In a study by Eisner et al. [7], overweight children were found to have a decreased amount of urinary oxalate. Such a finding may be associated with dietary factors in that patients with a higher BMI consume fewer oxalate-rich foods, such as nuts and vegetables; they may also consume larger amounts of fluid, which would account for increased urinary volume. In a study by Sarica et al. [20], pediatric patients considered to be obese based on their BMI were
Table 3 Nationwide Childrens Hospital patient characteristics and body composition results versus the 20032004 NHANES population Gender Nationwide Childrens Hospital Subgroup n %BF (mean SD) NHANES p value

%BF (mean SD) 0.1012a 0.6432a 0.0072*a 0.8774a 0.9759a 0.8421a 0.1772b 0.7622b 0.0143*b

Male (M) Female (F) M+F

All Kidney stones Hypercalciuria All Kidney stones Hypercalciuria All Kidney stones Hypercalciuria

24 11 13 26 18 8 50 29 21

25.211.2 303 26.68.5 29.29.6 21.811.6 32.88 315 33.47 33.18.8 32.26.4 29.210.3 618 308.2 30.35.5 25.711.1

*p <0.05 vs. NHANES (National Health and Nutrition Examination Survey)

a b

Controlled by age Controlled by age and gender

found to have higher urine pH values than the controls. In the same study, however, obese patients did have higher urinary oxalate excretion. It should be noted that the classifications of obesity for Sarica et al.s study was that all patients with a BMI 25 were considered obese and that controls were those with BMI <25. The authors did not take into account changes in BMI based on age and gender. In our current study, children with hypercalciuria and an increased %BF had a significantly higher occurrence of kidney stones compared to children with hypercalciuria and a normal %BF. No significant associations between body composition and 24-h urine chemistries were noted in the NCH population. This may have been due to the small sample size or the existence of other metabolic abnormalities predisposing obese children to form kidney stones. Adult studies have reported increased BMI values as a risk factor for increased urine calcium and sodium and decreased urine pH and volume [7, 21]. Published studies in children have noted an association with increased BMI and decreased urine pH and oxalate compared to normal controls [7, 19]. Such findings suggest that the relationship between body composition, urine chemistries, and stone formation may be unique to the pediatric population. In addition to known pathways of kidney stone formation, recent studies have examined the role of oxidative stress, tubular cell injury, inflammation, and fibrosis in kidney stone formation [22, 23]. Key modulators identified in this process, osteopontin and monocyte chemoattractant protein-1, are known to play a role in the development of insulin resistance, a main feature of the metabolic syndrome [24]. The OtsukaLongEvans Tokushima Fatty (OLETF) rat model has increased calcium oxalate kidney stone formation despite having similar urine pH, uric acid, calcium and oxalate parameters as the non-obese control rats [23]. The overall mean %BF in the NCH study group was not significantly different then the NHANES control group (29% NCH vs. 30% NHANES). Other pediatric population studies have yielded %BF measurements consistently lower

Pediatr Nephrol (2011) 26:21732178

2177 2. Lande MB, Varade W, Erkan E, Niederbracht Y, Schwartz GJ (2005) Role of urinary supersaturation in the evaluation of children with urolithiasis. Pediatr Nephrol 20:491494 3. Edvardsson V, Elidottir H, Indridason OS, Palsson R (2005) High incidence of kidney stones in Icelandic children. Pediatr Nephrol 20:940944 4. Sas DJ, Hulsey TC, Shatat IF, Orak JK (2010) Increasing incidence of kidney stones in children evaluated in the emergency department. J Pediatr 157:132137 5. Ebbeling CB, Pawlak DB, Ludwig DS (2002) Childhood obesity: public-health crisis, common sense cure. Lancet 360:473482 6. Duffey BG, Pedro RN, Kriedberg C, Weiland D, Melquist J, Ikramuddin S, Kellogg T, Makhlouf AA, Monga M (2008) Lithogenic risk factors in the morbidly obese population. J Urol 179:14011406 7. Eisner BH, Eisenberg ML, Stoller ML (2009) Influence of body mass index on quantitative 24-hour urine chemistry studies in children with nephrolithiasis. J Urol 182:11421145 8. Mei Z, Grummer-Strawn LM, Wang J, Thornton JC, Freedman DS, Pierson RN Jr, Dietz WH, Horlick M (2007) Do skinfold measurements provide additional information to body mass index in the assessment of body fatness among children and adolescents? Pediatrics 119:e1306e1313 9. Koyun M, Guven AG, Filiz S, Akman S, Akbas H, Baysal YE, Dedeoglu N (2007) Screening for hypercalciuria in schoolchildren: what should be the criteria for diagnosis? Pediatr Nephrol 22:12971301 10. Ghazali S, Barratt TM (1974) Urinary excretion of calcium and magnesium in children. Arch Dis Child 49:97101 11. Flegal KM, Ogden CL, Yanovski JA, Freedman DS, Shepherd JA, Graubard BI, Borrud LG (2010) High adiposity and high body mass index-for-age in US children and adolescents overall and by race-ethnic group. Am J Clin Nutr 91:10201026 12. Shypailo RJ, Butte NF, Ellis KJ (2008) DXA: can it be used as a criterion reference for body fat measurements in children? Obesity (Silver Spring) 16:457462 13. Higgins PB, Gower BA, Hunter GR, Goran MI (2001) Defining health-related obesity in prepubertal children. Obes Res 9:233 240 14. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R, Mei Z, Curtin LR, Roche AF, Johnson CL (2000) CDC growth charts: United States. Adv Data 8:127 15. Kuczmarski RJ, Flegal KM (2000) Criteria for definition of overweight in transition: background and recommendations for the United States. Am J Clin Nutr 72:10741081 16. Li WM, Chou YH, Li CC, Liu CC, Huang SP, Wu WJ, Chen CW, Su CY, Lee MH, Wei YC, Huang CH (2009) Association of body mass index and urine pH in patients with urolithiasis. Urol Res 37:193196 17. Daudon M, Lacour B, Jungers P (2006) Influence of body size on urinary stone composition in men and women. Urol Res 34:193199 18. Ekeruo WO, Tan YH, Young MD, Dahm P, Maloney ME, Mathias BJ, Albala DM, Preminger GM (2004) Metabolic risk factors and the impact of medical therapy on the management of nephrolithiasis in obese patients. J Urol 172:159163 19. Sarica K (2006) Pediatric urolithiasis: etiology, specific pathogenesis and medical treatment. Urol Res 34:96101 20. Sarica K, Eryildirim B, Yencilek F, Kuyumcuoglu U (2009) Role of overweight status on stone-forming risk factors in children: a prospective study. Urology 73:10031007 21. Siener R, Glatz S, Nicolay C, Hesse A (2004) The role of overweight and obesity in calcium oxalate stone formation. Obes Res 12:106113 22. Khan SR (2004) Crystal-induced inflammation of the kidneys: results from human studies, animal models, and tissue-culture studies. Clin Exp Nephrol 8:7588

than both the NCH hypercalciuria population and the 2003 2004 NHANES population [12, 25, 26], raising the possibility that the %BF was higher in hypercalciuric children than controls. It is more likely, however, that the higher %BF in the NCH hypercalciuria and the control population reflects the increasing rate of pediatric obesity in the US. There are limitations to our study that merit consideration. Our study was a retrospective chart review and therefore subject to the inherent problems of a non-prospective design. Recent studies have used DXA-derived %BF measurements as a criterion or reference for comparison with other body composition measurement techniques [7, 27]. Despite the increased use of DXA, there are as yet no generally accepted %BF cutoffs for overweight and obesity in children and adolescents based on DXA measurements. %BF curves have been generated for body fat based on bio-impedance measurements [25]. The use of these bio-impedance curves in the NCH study population would have resulted in a mean %BF of 85% for age in females and 91% for age in males. Prospective studies in adults have shown that obesity and weight gain increase the risk of kidney stone formation in both males and females. In most adult studies, BMI is used to classify adult overweight and obesity based on healthrelated outcomes. However, the use of BMI in the pediatric population as a measure of adiposity is limited due to the BMI varying according to age, sex, and maturation. An additional limitation, for all age groups, is that the relative risk of health-related outcomes associated with certain BMI values is population-dependent [28]. We found that an increased %BF was associated with an increased occurrence of kidney stones in children with idiopathic hypercalciuria. The 24-h urine chemistries analyzed for stone risk factors in these patients, however, were not significantly different from those patients with a normal %BF. Interestingly, classic stone risk factors, such as urine pH and urine citrate, did not change with increased %BF. Because kidney stones have a tendency to recur and cause significant morbidity, research focused on providing a better understanding of pediatric kidney stone pathology is imperative. Future studies aimed at identifying common, modifiable risk factors in the development of kidney stones in the pediatric population will guide new approaches to the treatment and prevention of disease.
Acknowledgments Dr. Schwaderer is supported by a grant from the National Institute of Health (IRC40K090937-01).

References
1. Alon US, Zimmerman H, Alon M (2004) Evaluation and treatment of pediatric idiopathic urolithiasis-revisited. Pediatr Nephrol 19:516520

2178 23. Okamoto M, Kohjimoto Y, Iba A, Saji F, Hara I, Shigematsu T (2010) Calcium oxalate crystal deposition in metabolic syndrome model rat kidneys. Int J Urol 17:9961003 24. Tilg H, Moschen AR (2008) Inflammatory mechanisms in the regulation of insulin resistance. Mol Med 14:222231 25. McCarthy HD, Cole TJ, Fry T, Jebb SA, Prentice AM (2006) Body fat reference curves for children. Int J Obes (Lond) 30:598 602 26. Sopher AB, Thornton JC, Wang J, Pierson RN Jr, Heymsfield SB, Horlick M (2004) Measurement of percentage of body fat in 411

Pediatr Nephrol (2011) 26:21732178 children and adolescents: a comparison of dual-energy X-ray absorptiometry with a four-compartment model. Pediatrics 113:12851290 27. Elberg J, McDuffie JR, Sebring NG, Salaita C, Keil M, Robotham D, Reynolds JC, Yanovski JA (2004) Comparison of methods to assess change in children's body composition. Am J Clin Nutr 80:6469 28. Guo SS, Roche AF, Chumlea WC, Gardner JD, Siervogel RM (1994) The predictive value of childhood body mass index values for overweight at age 35 y. Am J Clin Nutr 59:810819