Abstract
Background
Obesity is prevalent among children with chronic kidney disease (CKD) and is associated with cardiovascular disease and reduced quality of life. Its relationship with pediatric CKD progression has not been described.
Methods
We evaluated relationships between both body mass index (BMI) category (normal, overweight, obese) and BMI z-score (BMIz) change on CKD progression among participants of the Chronic Kidney Disease in Children study. Kaplan–Meier survival curves and multivariable parametric failure time models depict the association of baseline BMI category on time to kidney replacement therapy (KRT). Additionally, the annualized percentage change in estimated glomerular filtration rate (eGFR) was modeled against concurrent change in BMIz using multivariable linear regression with generalized estimating equations which allowed for quantification of the effect of BMIz change on annualized eGFR change.
Results
Participants had median age of 10.9 years [IQR: 6.5, 14.6], median eGFR of 50 ml/1.73 m2 [IQR: 37, 64] and 63% were male. 160 (27%) of 600 children with non-glomerular and 77 (31%) of 247 children with glomerular CKD progressed to KRT over a median of 5 years [IQR: 2, 8]. Times to KRT were not significantly associated with baseline BMI category. Children with non-glomerular CKD who were obese experienced significant improvement in eGFR (+ 0.62%; 95% CI: + 0.17%, + 1.08%) for every 0.1 standard deviation concurrent decrease in BMI. In participants with glomerular CKD who were obese, BMIz change was not significantly associated with annualized eGFR change.
Conclusion
Obesity may represent a target of intervention to improve kidney function in children with non-glomerular CKD.
Keywords: Obesity, Kidney failure, Pediatrics
Introduction
Obesity in pediatric chronic kidney disease (CKD) is an increasingly prevalent problem, reported in up to 22% of children [1-3]. This is concerning given the evidence that children with CKD and obesity exhibit worse markers of cardiovascular disease [3-5], more depressive symptoms, and a reduced quality of life [6, 7]. Additionally, multiple large studies have identified an overweight or obese BMI at the time of pediatric transplant to be a significant risk factor for poorer transplant outcomes including graft failure [8-10]. A study of Israeli adolescents identified that obesity and overweight status at 17 years of age were associated with markedly elevated risks for future kidney failure [11]. Likewise, in the general adult population, obesity is associated with the development of CKD [12-15].
The relationship between overweight/obesity and CKD progression in those with CKD is more complex. In adults with preexisting CKD, obesity was not associated with progression of disease or worse clinical outcomes [16, 17]. In a study of adults with CKD, body mass index (BMI) > 35 kg/m2 was associated with worse clinical outcomes, though the best clinical outcomes were observed in patients who were overweight or mildly obese [16]. A study by Lalan et al. identified that children with CKD and multiple metabolic risk factors experienced a faster decline of kidney function, but did not specifically evaluate the role of obesity in kidney function change [3]. Accordingly, the relationships of obesity and changes in BMI with disease progression in children with CKD are unknown but important to elucidate in order to identify opportunities for targeted interventions to slow disease progression in children with CKD. In this analysis, we hypothesized that higher BMI would associate with faster progression of disease, i.e., shorter times to KRT and steeper GFR decline, and a reduction in BMI would associate with slower progression of disease, specifically less GFR decline.
Methods
Study population
The Chronic Kidney Disease in Children (CKiD) study is a multicenter, prospective cohort study of children and adolescents with mild-to-moderate CKD across North America. An observational study-monitoring board appointed by the National Institute of Diabetes and Digestive and Kidney Diseases and the institutional review boards of each participating center approved the study design and conduct with adherence to the Declaration of Helsinki. Participants and parents/legal guardians provided informed assent and consent. The design and objectives of the study have been previously published [18]. Since 2005, participants have been seen at annual follow-up visits and provide data on renal, cardiovascular, neurocognitive, and growth parameters.
Primary outcomes
We studied two primary outcomes of CKD progression. First, we defined disease progression as the time from the baseline CKiD visit to the initiation of kidney replacement therapy (KRT; dialysis or transplant). Dates of KRT were obtained by participant interview and/or clinical chart review. Participants were censored at 6 months after their last visit if they did not initiate KRT. Participants who died prior to initiation of KRT were censored at their last follow-up visit. Secondly, we quantified disease progression as the annualized percentage change in estimated glomerular filtration rate (eGFR) within each pair of consecutive visits throughout follow-up. Specifically, annualized percentage change in eGFR within two consecutive visits i and i + 1 was calculated by: (exp([loge(eGFRi+1/eGFRi)] / time in years between visits i and i + 1) − 1) × 100%. GFR was estimated at each annual visit by the CKiD-derived estimating equation using serum creatinine and cystatin C concentrations [19]. Serum creatine, assayed enzymatically, and cystatin C, assayed nephelometrically, were measured at the CKiD central biochemistry laboratory at each annual visit and calibrated to International Federation of Clinical Chemistry standards.
Primary exposure
Age–sex-specific BMI z-scores (BMIz) and BMI categories (normal [5th to < 85th percentile], overweight [≥ 85th to < 95th percentile], or obese [≥ 95th percentile]) were determined at each visit in which participants were than less than 20 years of age based on the 2000 US Centers for Disease Control and Prevention (CDC) standard growth charts. BMI was classified according to CDC standard growth charts, instead of height age, as proposed by Gao et al., since use of height age does not account for age-specific dynamics between linear size and body composition that are a function of both pubertal maturation and time (i.e., age) [20]. Use of the CDC charts avoids inappropriately comparing a short pubertal child with an average height pre-pubertal child, which leads to misclassification of healthy weight children as overweight/obese. Participants whose BMI was characterized as underweight (< 5th percentile) at baseline or at any visit during follow-up were excluded entirely from analyses.
Covariates
Demographic covariates include sex (male versus female) and age (years). Clinical covariates include eGFR, blood pressure (BP), and proteinuria, since these measures are known to associate with CKD progression [21]. BP measurements are obtained by auscultation at study enrollment, at a follow-up visit 6 months later, and annually thereafter. The CKiD Clinical Coordinating Centers provide all sites with the same aneroid sphygmomanometer (Mabis MedicKit 5; Mabis Healthcare, Waukegan, IL). Annually, CKiD clinical staff are trained and certified in the auscultatory BP measurement technique, and each center’s aneroid device is calibrated. At each visit, three BP measurements at 30-s intervals are obtained by auscultation of the brachial artery using the first Korotkoff sound for systolic BP and the fifth Korotkoff sound for diastolic BP. The average of the three BP measurements is recorded as the participant’s casual BP for that visit. Casual systolic and diastolic BP measurements were standardized (z-scores and percentiles) for age, sex, and height using the American Academy of Pediatrics 2017 updated BP guidelines based on normal-weight children [22]. Specifically, we used three BP categories in analyses: normal BP, elevated BP, and Stage 1 Hypertension or Stage 2 Hypertension. Participants provided a random urine collection on the morning of a study visit. The total urine protein content was determined at the CKiD central biochemistry laboratory by a turbidimetric method using benzethonium chloride (Roche Diagnostics) which reacts with urine protein in a basic medium to produce a turbidity. The quantity of protein (mg/dL) was divided by the quantity of creatinine (mg/dL) in the urine to obtain the urine protein:creatinine ratio (UPC). Proteinuria was categorized by level of UPC < 0.5, 0.5–2, and > 2. Corticosteroid use was also included as a covariate in analyses for participants with glomerular CKD only. We did not include race, ethnicity, or socioeconomic level as covariates in our analyses since health disparities were not outcomes or exposures of consideration for this study. Further supporting the exclusion of socioeconomic covariates, we have previously shown that race, maternal education, and household income were not associated with BMIz change over time in this population [2].
Statistical analyses
In the first set of analyses, participants were stratified into one of three groups based on their baseline BMI categorization (normal, overweight, obese). Kaplan–Meier survival curves were used to depict the distribution of times from baseline visit to KRT. Parametric failure time models assuming a generalized gamma GG distribution (with location parameter β, scale parameter σ, and shape parameter κ) of times from baseline visit to KRT initiation were fit to the three Kaplan–Meier survival curves and used to estimate relative times (RT) which summarize the impact a baseline BMI category of overweight or obese (relative to normal BMI) has on the time to KRT [23]. Specifically, an RT represents the time it takes for P% of the overweight or obese group to initiate KRT divided by the time it takes for the same P% of the normal BMI group to initiate KRT. We chose to use parametric models as opposed to semi-parametric Cox models because we wanted to use the RT as opposed to the relative hazard as the measure of association between the exposure and the outcome, since time (years) was the unit of measurement of the dependent variable. We utilized regression models under the assumption of proportional times (i.e., the RT did not depend on “P”) because the more complex models which allow for non-proportional times did not provide a better fit to the data. The RT for the overweight and obese baseline BMI categories compared to the normal baseline BMI category are reported, unadjusted, and adjusted for the previously defined covariates measured at baseline.
In the second set of analyses, a pair of sequential visits i and i + 1 that were required to occur between 0.5 and 2.5 years of each other were the unit of analysis. The initial BMI category was defined as the BMI category at the first visit within the visit pair (at visit i). The annualized change in BMIz within two consecutive visits i and i + 1 was calculated as (BMIzi+1 − BMIzi) / time in years between visits i and i + 1. The annualized percentage change in eGFR was calculated for each visit pair (defined above) and regressed on the annualized change in BMIz; the mean annualized change in eGFR was estimated for the reference (no change in BMIz between the two visits) by the intercept of the regression model and the effect of a one-tenth of one standard deviation change in annualized BMI on the annualized eGFR was quantified by the product of 0.10 and the slope of the regression model. Three separate regression models were used (each model corresponded to one of the three initial BMI categories) and generalized estimating equations were used to account for repeated visit pairs contributed by the same participant. While there was heterogeneity in the number of visit pairs contributed to the analysis by participants within a given BMI category, there was no evidence that initial BMIz was associated with the number of visit pairs contributed to the analyses. Unadjusted and estimates adjusted for previously defined covariates measured at the initial visit (at visit i) of the visit pair were reported. Each regression model included values of time updated covariates (e.g., proteinuria, BP category) at the initial visit (at visit i) of the visit pair, since current values at visit i are likely to confound the relationship between BMIz change and eGFR change to a greater extent than values from the baseline visit. In all analyses, 95% confidence intervals were used as a measure of precision. CKD progression and characteristics related to CKD progression have been shown to be different in children with non-glomerular and glomerular causes of CKD [24-26], therefore, all analyses were done separately by CKD etiology.
Results
As of November 2020, 1079 participants were enrolled in the CKiD study. There were 1023 (95%) participants enrolled who had both eGFR and BMIz available: 3% underweight, 65% normal weight, 15% overweight, and 16% obese at baseline. We excluded 88 participants who were underweight at any point during the study since the number of underweight participants was too small to make meaningful conclusions and were out of the scope of this paper. Our time-to-event analyses included 847 participants with complete baseline exposure data and all covariates of interest, as well as follow-up time at risk for KRT: 600 participants with non-glomerular CKD and 247 participants with glomerular CKD (Fig. 1). For those with non-glomerular disease, 423 (71%) had a normal BMI, 85 (14%) had an overweight BMI, and 92 (15%) had an obese BMI at baseline. For those with glomerular disease, 134 (54%), 49 (20%), and 64 (26%) had normal BMI, overweight BMI, and obese BMI at baseline, respectively. Baseline characteristics of the 600 participants with non-glomerular CKD etiology and the 247 participants with glomerular CKD are shown in Table 1 by baseline BMI category.
Fig. 1.
Chronic Kidney Disease in Children (CKiD) study participants and inclusion criteria for analyses
Table 1.
Baseline demographic and clinical characteristics for 600 children with non-glomerular chronic kidney disease (CKD) and 247 children with glomerular CKD stratified into three baseline body mass index (BMI) categories.a
| Non-Glomerular (NG) CKD, n = 600 |
Glomerular (G) CKD, n = 247 |
|||||
|---|---|---|---|---|---|---|
| Characteristics b | Normal (n = 423) | Overweight (n = 85) | Obese (n = 92) |
Normal (n = 134) | Overweight (n = 49) | Obese (n = 64) |
| CKD duration, years | 8.2 [4.5, 12.9] | 10.3 [5.0, 12.6] | 9.0 [5.3, 11.4] | 4.2 [1.8, 8.5] | 2.3 [0.9, 6.0] | 2.9 [1.1, 4.7] |
| Age, years | 8.9 [4.8, 13.3] | 11.0 [5.4, 13.7] | 9.3 [6.5, 12.9] | 14.2 [10.1, 15.9] | 14.0 [11.0, 15.9] | 14.3 [11.8, 16.0] |
| Male Sex | 67% (283) | 66% (56) | 71% (65) | 54% (72) | 47% (23) | 55% (35) |
| Race | ||||||
| White | 71% (300) | 71% (60) | 67% (62) | 59% (78) | 47% (23) | 41% (26) |
| Black | 18% (75) | 19% (16) | 24% (22) | 25% (34) | 37% (18) | 39% (25) |
| Other | 11% (47) | 10% (9) | 9% (8) | 16% (22) | 16% (8) | 20% (13) |
| Maternal education | ||||||
| High school | 35% (145) | 33% (28) | 43% (39) | 41% (54) | 45% (21) | 57% (35) |
| Some College | 30% (125) | 29% (24) | 24% (22) | 22% (29) | 20% (9) | 25% (15) |
| College Graduate | 35% (145) | 38% (32) | 32% (29) | 37% (49) | 35% (16) | 18% (11) |
| Proteinuria | ||||||
| < 0.5 | 66% (280) | 57% (48) | 71% (65) | 50% (67) | 45% (22) | 31% (20) |
| 0.5 to 2.0 | 27% (115) | 28% (24) | 23% (21) | 27% (36) | 31% (15) | 42% (27) |
| > 2.0 | 7% (28) | 15% (13) | 7% (6) | 23% (31) | 24% (12) | 27% (17) |
| eGFR c | ||||||
| < 30 | 17% (73) | 7% (6) | 11% (10) | 10% (14) | 4% (2) | 6% (4) |
| ≥ 30 to 45 | 31% (129) | 24% (20) | 27% (25) | 20% (27) | 24% (12) | 22% (14) |
| ≥ 45 to 60 | 28% (119) | 42% (36) | 29% (27) | 24% (32) | 22% (11) | 28% (18) |
| ≥ 60 | 24% (102) | 27% (23) | 33% (30) | 46% (61) | 50% (24) | 44% (28) |
| Blood pressure d | ||||||
| Normal | 60% (254) | 54% (46) | 54% (50) | 70% (94) | 51% (25) | 47% (30) |
| Elevated BP | 12% (51) | 16% (14) | 12% (11) | 12% (16) | 14% (7) | 28% (18) |
| Hypertension e | 28% (118) | 30% (25) | 34% (31) | 18% (24) | 35% (17) | 25% (16) |
| Albumin, g/dL | 4.4 [4.2, 4.6] | 4.5 [4.3, 4.7] | 4.4 [4.2, 4.6] | 4.2 [3.8, 4.5] | 4.0 [3.7, 4.3] | 4.1 [3.4, 4.4] |
| Height z-score | −0.66 [−1.40, 0.17] | −0.42 [−1.32, 0.14] | −0.10 [−0.79, 0.63] | −0.46 [−1.06, 0.27] | 0.00 [−0.75, 0.76] | 0.24 [−0.39, 1.24] |
| Corticosteroid use | 1% (4) | 0% (0) | 3% (3) | 22% (30) | 41% (20) | 41% (26) |
At baseline all participants are ≤ 16 years. Normal category defined as ≥ 5th to < 85th BMI percentile; Overweight defined as ≥ 85th to < 95th BMI percentile; Obese defined as ≥ 95th BMI percentile
Missing data includes CKD duration: n = 12 (4 NG + 8 G); race: n = 1 NG; maternal education: n = 19 (11 NG + 8 G); albumin: n = 7 (4 NG + 3 G); height z-score: n = 6 (4 NG + 2 G)
Estimated glomerular filtration rate (eGFR) calculated from averaging equations provided in Kidney International 2021; 99:948–956
Blood pressure (BP) categories determined by American Academy of Pediatrics (AAP) 2017 updated BP guidelines
Hypertension defined as either Stage 1 or Stage 2 hypertension based on the AAP 2017 guidelines
Progression to KRT occurred in 160 (27%) children with non-glomerular disease and 77 (31%) children with glomerular disease over a median follow-up time of 5 years [IQR: 2, 8] (Fig. 2 and Table 2). Five children died during the study and were censored at the date of their last follow-up visit. In the non-glomerular group, the percentage of participants who progressed to KRT by baseline BMI category was 27% of those with normal weight, 21% of those who were overweight, and 29% of those who were obese. For those with glomerular disease, 32%, 34%, and 28% of the participants with normal weight, overweight, and obesity progressed to KRT, respectively. There were no significant differences in the distribution of time to KRT by baseline BMI in either the non-glomerular or glomerular groups. As suggested by the Kaplan–Meier curves (Fig. 2) and after adjustment for sex, age, eGFR, proteinuria, and hypertension status, overweight participants with non-glomerular CKD had 26% longer time to KRT than participants with normal BMI; however, the difference was not statistically significant (Table 2). Participants with non-glomerular CKD who were obese at baseline had similar time to KRT as participants with normal BMI. In the glomerular CKD group, even after adjustment, each of the three baseline BMI groups had similar time to KRT.
Fig. 2.
Kaplan–Meier survival curves (solid lines) and appropriateness of Generalized Gamma (GG) parametric models (dashed lines) fit to the data of time from baseline Chronic Kidney Disease in Children (CKiD) visit to kidney replacement therapy (KRT) initiation stratified by baseline body mass index (BMI) category are shown for 600 children with non-glomerular CKD (left panel) and 247 children with glomerular CKD (right panel). The location, scale, and shape parameter estimates of the GG models for the three BMI categories of children with non-glomerular CKD were: normal BMI (2.58, 0.80, 0.60), overweight (2.79, 0.80, 0.60), and obese (2.54, 0.80, 0.60). The parameter estimates of the GG models for the three BMI categories of children with glomerular CKD were: normal BMI (2.21, 2.01, −0.60), overweight (2.10, 2.01, −0.60), and obese (2.33, 2.01, −0.60)
Table 2.
Unadjusted and adjusted relative times (RT) and 95% confidence intervals (CI) as measures of association between baseline body mass index (BMI) category and time to kidney replacement therapy (KRT) initiation among 600 children with non-glomerular chronic kidney disease (CKD) diagnoses and 247 children with glomerular CKD diagnoses
| Non-glomerular CKD (N = 600; 160 events) |
Glomerular CKD (N = 247; 77 events) |
|||||
|---|---|---|---|---|---|---|
| Baseline BMI Category | n; number of Events | Unadjusted RT (95% CI) | Adjusted a RT (95% CI) | n; number of Events | Unadjusted RT (95% CI) | Adjusted a, b RT (95% CI) |
| Normal | 423; 116 | 1 (reference) | 1 (reference) | 134; 42 | 1 (reference) | 1 (reference) |
| Overweight | 85; 17 | 1.23 (0.88, 1.73) | 1.26 (0.95, 1.66) | 49; 17 | 0.90 (0.44, 1.84) | 0.87 (0.53, 1.41) |
| Obese | 92; 27 | 0.96 (0.72, 1.29) | 0.91 (0.73, 1.15) | 64; 18 | 1.13 (0.55, 2.31) | 1.09 (0.70, 1.69) |
Adjusted for sex and baseline age, eGFR, proteinuria, and hypertension
Additionally adjusted for baseline corticosteroid use
Next, we restricted the study population to participants with BMIz and eGFR data as well as complete covariate data for at least one pair of consecutive visits that occurred within 0.5 to 2.5 years (median 1.02 years [IQR: 0.96, 1.13]) (Fig. 1). This resulted in a total of 3067 visit pairs contributed by 721 participants: 2330 visit pairs from 505 participants with non-glomerular CKD and 737 visit pairs from 216 participants with glomerular CKD. The median (IQR) of the distribution of annualized BMIz changes in those with non-glomerular CKD were 0.02 (−0.24, 0.26), −0.03 (−0.25, 0.13), and −0.01 (−0.16, 0.10) for the normal, overweight, and obese groups, respectively. Fifty-three percent of the 391 visit pairs contributed by participants with non-glomerular CKD and obesity showed a decline in BMIz; the median annualized BMI decline was −0.15 SD (IQR: −0.30, −0.05) in these 207 visit pairs. For participants with glomerular CKD, median (IQR) changes in annualized BMIz were −0.03 (−0.28, 0.22), −0.04 (−0.26, 0.14), and −0.05 (−0.20, 0.06) for the normal, overweight, and obese groups, respectively. Of the participants with glomerular CKD who were obese, sixty percent of the 189 visit pairs experienced a decline in BMIz with a median annualized BMI of −0.16 SD (IQR: −0.33, −0.06) in the 113 visit pairs.
Table 3 displays the unadjusted and adjusted annualized percentage change in eGFR by CKD etiology and initial BMI category in each visit pair. Specifically, a participant with non-glomerular CKD, obesity, and reference values for the covariates without a change in BMIz within a visit pair has an average annualized eGFR change of −2.5% (95% CI: −6.1%, + 1.1%); similar to −2.7% (95% CI: −5.4%, + 0.02%) and −1.2% (95% CI: −2.5%, + 0.1%) for those who are overweight or normal weight at the initial visit, respectively. However, among participants with non-glomerular CKD who were obese, each 0.1 standard deviation decrease in annualized BMI was significantly associated with a concurrent + 0.62% increase (95% CI: + 0.17%, + 1.08%) in annualized eGFR. For those with glomerular CKD, there was no significant adjusted annualized change in eGFR in any of the weight categories, regardless of annualized change in BMIz.
Table 3.
Unadjusted and adjusted annualized percentage changes in estimated glomerular filtration rate (eGFR) by chronic kidney disease (CKD) etiology and initial body mass index (BMI) category in the visit pair (normal, overweight, obese)
| Non-Glomerular CKD |
Glomerular CKD |
|||||
|---|---|---|---|---|---|---|
| Normal1613 visit pairs (95% CI) |
Overweight 326 visit pairs (95% CI) |
Obese 391 visit pairs (95% CI) |
Normal 410 visit pairs (95% CI) |
Overweight 138 visit pairs (95% CI) |
Obese 189 visit pairs (95% CI) |
|
| Unadjusted Analyses | ||||||
| Mean Annualized eGFR change among constant BMIz | −2.5% (−3.2%, −1.8%) | −3.4% (−4.8%, −2.1%) | −4.4% (−5.9%, −2.8%) | −6.0% (−8.3%, −3.6%) | −4.4% (−7.7%, −1.1%) | −3.7% (−6.9%, −0.5%) |
| Mean effect of 0.10-SD decrease in BMI on annualized eGFR change | −0.01% (−0.26%, + 0.25%) | −0.19% (−0.55%, + 0.18%) | + 0.65% (+ 0.15%, + 1.15%) | −0.44% (−0.87%, −0.01%) | −0.05% (−1.33%, + 1.23%) | −0.27% (−1.21%, + 0.67%) |
| Adjusted Analysesa | ||||||
| Mean Annualized eGFR change among constant BMIz b | −1.2% (−2.5%, + 0.1%) | −2.7% (−5.4%, + 0.02%) | −2.5% (−6.1%, + 1.1%) | −3.9% (−8.8%, + 1.0%) | −0.6% (−5.4%, + 4.2%) | −4.0% (−8.8%, + 0.7%) |
| Mean effect of 0.10-SD decrease in BMI on annualized eGFR change | + 0.04% (−0.20%, + 0.27%) | −0.16% (−0.50%, + 0.17%) | + 0.62% (+ 0.17%, + 1.08%) | −0.32% (−0.76%, + 0.12%) | + 0.31% (−0.69%, + 1.31%) | −0.47% (−1.41%, + 0.47%) |
Each of the six analyses were adjusted for sex and first visit within the visit pair values of age, eGFR, proteinuria and hypertension. The three analyses in the BMI categories for Glomerular CKD were additionally adjusted for first visit within the visit pair values of corticosteroid use
The expected annualized eGFR change for a 15-year-old female without hypertension and with eGFR = 70 and proteinuria < 0.5 at the first visit within the visit pair and who does not change BMIz within the visit pair. For participants with glomerular CKD who did not use corticosteroids at the first visit within the visit pair
Discussion
This is the first study, of which we are aware, to quantify the relationship between BMI and time to KRT, as well as the relationship between BMI change and concomitant change in eGFR. We found a nonsignificant increase in time to KRT in the overweight group of children with non-glomerular CKD compared to children with either normal or obese weight and otherwise no effect of baseline BMI category on time to KRT. Consistent with Rodig et al. [2], we found that children in the cohort typically did not have much of a change in their BMIz between any two sequential visits. However, in 53% of visit pairs contributed by participants with non-glomerular CKD and obesity who had a decline in BMIz, the median BMI reduction was 0.15 standard deviations per year (IQR: −0.30, −0.05). For this group of children who had non-glomerular CKD and obesity, a BMIz reduction between two visits was associated with a concurrent significant increase in eGFR. Alternatively, those who maintained their BMIz within a pair of sequential visits (and had reference values for all covariates) experienced a nonsignificant decrease in annualized eGFR. These findings are especially notable given the expected reduction of hyperfiltration, and consequent increase in creatinine, that may occur with transition from an obese phenotype [27, 28]. The improvement in eGFR may be explained by reduction of inflammation, associated increases in physical activity, or improved metabolic parameters such as lipids and insulin sensitivity, all of which may impact kidney function and should be explored in future studies [5, 29-34].
In contrast to the non-glomerular CKD group, in the glomerular group we did not note any relationship between annualized percentage change in eGFR and change in BMIz after adjustment. The most pertinent of the null associations was in the participants with glomerular CKD who were obese. This finding however was in the opposite direction (decrease in BMIz was associated with a decrease in eGFR) as compared to the result found in participants with non-glomerular CKD who were obese (decrease in BMIz was associated with an increase in eGFR). Since the glomerular group in our study was smaller than the non-glomerular group, it may be of interest to confirm these results in another larger glomerular disease cohort. The null findings found in our study are consistent with results from studies evaluating progression of CKD based on BMI of adult patients with nephrotic syndrome. These studies similarly have not found BMI category to associate with progression of disease [17, 35]. It is possible that BMI in glomerular disease is confounded by edema and that true adiposity loss would yield similar results to the non-glomerular group. Supporting this hypothesis, in the glomerular disease group, those with a normal BMI were less likely to have nephrotic range proteinuria and low albumin than those who were obese. Additionally, glomerular disease is associated with faster progression of kidney disease, and this effect may be stronger than and override the effect of BMI on progression in this population [36, 37].
Although healthy adolescents with obesity are at an increased risk for future need for KRT [11], our finding of a lack of association between obesity and time to KRT is in congruence with the adult literature [16, 38, 39]. Lin et al., did not find an increased risk for KRT when evaluating obesity by both BMI and percent body fat using dual X-ray absorptiometry (DXA) in over 300 adults with CKD stages 3–5 [39]. Additionally, some reports have identified a reduced risk of KRT5 with higher BMI [16, 40, 41]. Research by Jun et al. of 453,946 veterans did not show a higher risk of need for KRT in obese patients with CKD, but found a reduced risk in the overweight group [16], as did a meta-analysis by Ahmadi et al. [40]. The possible protective effect of overweight BMI on progression of disease may explain why one large adult study did find a significant association with CKD progression and obesity when the reference group combined individuals with overweight and normal BMI [42]. It is not clear why overweight BMI may be protective against progression of kidney disease, but it may indicate better nutrition and lean muscle mass, which associate with better outcomes, and are more likely to be deficient in those who have a normal weight [43, 44]. Similarly, we noted a decreased risk of KRT in our overweight population with non-glomerular CKD, although after full adjustment the finding was not statistically significant. Although access to transplant and dialysis may be delayed in children with obesity or overweight, we would expect therefore, to see not only prolonged survival in the participants who are overweight, but also, and even more so, in the participants who are obese [45]. The absence of prolonged time to KRT in obese participants suggests that differential access to transplant is not confounding the results of our analysis.
The limitations of this study include our inability to fully assess other potentially important variables, including dietary intake, lean muscle mass, Tanner stage, and physical function. Additionally, we use BMI as a sole marker of adiposity even though there is data to suggest that when evaluated in conjunction with measures of lean mass or fat mass it offers a better prediction of outcomes [39, 46, 47]. It is possible that if we had additional measures of fat mass or lean mass, such as waist circumference, arm skinfold thickness, or DXA data, we would have identified stronger relationships between BMI and CKD progression, but unfortunately in the CKiD data set most participants do not have corresponding waist circumference or arm skinfold thickness, and none have DXA data. We also acknowledge that although BP category and proteinuria (and all time-varying covariates) were those at the visit immediately prior to eGFR, it is possible that these could have changed at some point after the initial visit (during the 0.5 to 2.5 years) and before the subsequent visit, and thus any change within a pair of visits is not captured by using a single measurement. We did explore whether those who were included in the analyses were different from those who were excluded from the analyses with respect to progressing to KRT; 28% of the 847 participants included in our analyses progressed to KRT, which was similar to 22% of the 232 participants who were excluded and progressed to KRT. This suggests that those CKiD participants who were not included were similar to those who were with respect to the occurrence (but not necessarily the timing) of the KRT outcome of interest. As with all observational studies, our findings are subject to possible unmeasured confounding. Strengths of this paper include the large sample size for a pediatric CKD population and the long-term follow-up. Lastly, we utilized validated combined cystatin C and creatinine equations to estimate GFR, providing for eGFR measurements that are less dependent on creatinine, which can be affected by muscle mass. These updated equations have been shown to provide accurate estimation of GFR in young children and young adults 18 years and older [19].
In conclusion, our study identifies that although obesity at baseline does not associate with faster CKD progression, BMIz reduction in children with non-glomerular CKD who are obese can improve kidney function to slow the progression of disease. In addition to the well-described cardiovascular and psychosocial effects of obesity in CKD and the negative impact on outcomes of obesity at the time of kidney transplant, these findings provide further evidence that lifestyle management to address obesity in the pre-CKD 5 period is of paramount importance.
Supplementary Material
Acknowledgements
Data in this manuscript were collected by the Chronic Kidney Disease in children prospective cohort study (CKiD) with clinical coordinating centers (Principal Investigators) at Children's Mercy Hospital and the University of Missouri—Kansas City (Bradley Warady, MD) and Children's Hospital of Philadelphia (Susan Furth, MD, PhD), Central Biochemistry Laboratory (George Schwartz, MD) at the University of Rochester Medical Center, and data coordinating center (Alvaro Muñoz, PhD and Derek Ng, PhD) at the Johns Hopkins Bloomberg School of Public Health. The CKiD website is located at https://statepi.jhsph.edu/ckid and a list of CKiD collaborators can be found at https://statepi.jhsph.edu/ckid/site-investigators/. Please refer to the supplemental document which contains a list of the site principal investigators. This manuscript was presented in abstract form as oral platform presentation at the Pediatric Academic Society Annual Meeting 2021.
Funding
The CKiD Study is funded by the National Institutes of Health (NIH)/National Institute of Diabetes and Digestive and Kidney Diseases with additional funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Heart, Lung, and Blood Institute (U01-DK-66143, U01-DK-66174, U24-DK-082194, and U24-DK66116). None of the funding sources had a role in study design; collection, analysis, and interpretation of data; writing the report; or the decision to submit the report for publication.
Abbreviations
- BMI
body mass index
- CKD
Chronic kidney disease
- KRT
Kidney replacement therapy
- CKiD
The Chronic Kidney Disease in Children study
- eGFR
Estimated glomerular filtration rate
- CDC
Centers for Disease Control and Prevention
- BP
Blood pressure
- RT
Relative times
Footnotes
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00467-022-05655-6.
Financial disclosure The authors declare that they have no relevant financial interests.
References
- 1.Filler G, Reimão SM, Kathiravelu A, Grimmer J et al. (2007) Pediatric nephrology patients are overweight: 20 years’ experience in a single Canadian tertiary pediatric nephrology clinic. Int Urol Nephrol 39:1235–1240. 10.1007/s11255-007-9258-y [DOI] [PubMed] [Google Scholar]
- 2.Rodig NM, Roem J, Schneider MF, Seo-Mayer PW et al. (2021) Longitudinal outcomes of body mass index in overweight and obese children with chronic kidney disease. Pediatr Nephrol 36:1851–1860. 10.1007/s00467-020-04907-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lalan S, Jiang S, Ng DK, Kupferman F et al. (2018) Cardiometabolic Risk Factors, Metabolic Syndrome, and Chronic Kidney Disease Progression in Children. J Pediatr 202:163–170. 10.1016/j.jpeds.2018.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Schaefer F, Doyon A, Azukaitis K, Bayazit A et al. (2017) Cardiovascular Phenotypes in Children with CKD: The 4C Study. Clin J Am Soc Nephrol 12:19–28. 10.2215/CJN.01090216 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wilson AC, Schneider MF, Cox C, Greenbaum LA et al. (2011) Prevalence and Correlates of Multiple Cardiovascular Risk Factors in Children with Chronic Kidney Disease. Clin J Am Soc Nephrol 6:2759–2765. 10.2215/CJN.03010311 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kogon AJ, Matheson MB, Flynn JT, Gerson AC et al. (2016) Depressive Symptoms in Children with Chronic Kidney Disease. J Pediatr 168:164–170.e1. 10.1016/j.jpeds.2015.09.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kogon AJ, Kim JY, Laney N, Radcliffe J et al. (2019) Depression and neurocognitive dysfunction in pediatric and young adult chronic kidney disease. Pediatr Nephrol 34:1575–1582. 10.1007/s00467-019-04265-z [DOI] [PubMed] [Google Scholar]
- 8.Ladhani M, Lade S, Alexander SI, Baur LA et al. (2017) Obesity in pediatric kidney transplant recipients and the risks of acute rejection, graft loss and death. Pediatr Nephrol 32:1443–1450. 10.1007/s00467-017-3636-1 [DOI] [PubMed] [Google Scholar]
- 9.Kaur K, Jun D, Grodstein E, Singer P et al. (2018) Outcomes of underweight, overweight, and obese pediatric kidney transplant recipients. Pediatr Nephrol 33:2353–2362. 10.1007/s00467-018-4038-8 [DOI] [PubMed] [Google Scholar]
- 10.Winnicki E, Dharmar M, Tancredi DJ, Nguyen S et al. (2018) Effect of BMI on allograft function and survival in pediatric renal transplant recipients. Pediatr Nephrol 33:1429–1435. 10.1007/s00467-018-3942-2 [DOI] [PubMed] [Google Scholar]
- 11.Vivante A, Golan E, Tzur D, Leiba A et al. (2012) Body mass index in 1.2 million adolescents and risk for end-stage renal disease. Arch Intern Med 172:1644–1650. 10.1001/2013.jamainternmed.85 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Garofalo C, Borrelli S, Minutolo R, Chiodini P et al. (2017) A systematic review and meta-analysis suggests obesity predicts onset of chronic kidney disease in the general population. Kidney Int 91:1224–1235. 10.1016/j.kint.2016.12.013 [DOI] [PubMed] [Google Scholar]
- 13.Cao X, Zhou J, Yuan H, Wu L et al. (2015) Chronic kidney disease among overweight and obesity with and without metabolic syndrome in an urban Chinese cohort. BMC Nephrol 16:85. 10.1186/s12882-015-0083-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ejerblad E, Fored CM, Lindblad P, Fryzek J et al. (2006) Obesity and risk for chronic renal failure. J Am Soc Nephrol 17:1695–1702. 10.1681/ASN.2005060638 [DOI] [PubMed] [Google Scholar]
- 15.Hsu CY, McCulloch CE, Iribarren C, Darbinian J et al. (2006) Body mass index and risk for end-stage renal disease. Ann Intern Med 144:21–28. 10.7326/0003-4819-144-1-200601030-00006 [DOI] [PubMed] [Google Scholar]
- 16.Jun LL, Kamyar KZ, Jennie Z, Darryl Quarles L et al. (2014) Association of body mass index with outcomes in patients with CKD. J Am Soc Nephrol 25:2088–2096. 10.1681/ASN.2013070754 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Elyan BMP, Lees JS, Gillis KA, Mackinnon B et al. (2019) Obesity is not associated with progression to end stage renal disease in patients with biopsy-proven glomerular diseases. BMC Nephrol 20:237. 10.1186/s12882-019-1434-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Furth SL, Cole SR, Moxey-Mims M, Kaskel F et al. (2006) Design and methods of the Chronic Kidney Disease in Children (CKiD) prospective cohort study. Clin J Am Soc Nephrol 1:1006–1015. 10.2215/CJN.01941205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pierce CB, Muñoz A, Ng DK, Warady BA et al. (2021) Age- and sex-dependent clinical equations to estimate glomerular filtration rates in children and young adults with chronic kidney disease. Kidney Int 99:948–956. 10.1016/j.kint.2020.10.047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gao T, Leonard MB, Zemel B, Kalkwarf HJ et al. (2012) Interpretation of body mass index in children with CKD. Clin J Am Soc Nephrol 7:558–564. 10.2215/CJN.09710911 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Warady BA, Abraham AG, Schwartz GJ, Wong CS et al. (2015) Predictors of Rapid Progression of Glomerular and Nonglomerular Kidney Disease in Children and Adolescents: The Chronic Kidney Disease in Children (CKiD) Cohort. Am J Kidney Dis 65:878–888. 10.1053/J.AJKD.2015.01.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Flynn JT, Kaelber DC, Baker-Smith CM, Blowey D et al. (2017) Clinical practice guideline for screening and management of high blood pressure in children and adolescents. Pediatrics 140:e20171904. [DOI] [PubMed] [Google Scholar]
- 23.Cox C, Chu H, Schneider MF, Muñoz A (2007) Parametric survival analysis and taxonomy of hazard functions for the generalized gamma distribution. Stat Med 26:4352–4374. 10.1002/sim.2836 [DOI] [PubMed] [Google Scholar]
- 24.Ng DK, Portale AA, Furth SL, Warady BA et al. (2018) Time-varying coefficient of determination to quantify the explanatory power of biomarkers on longitudinal GFR among children with chronic kidney disease. Ann Epidemiol 28:549–556. 10.1016/j.annepidem.2018.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pierce CB, Cox C, Saland JM, Furth SL et al. (2011) Methods for characterizing differences in longitudinal glomerular filtration rate changes between children with glomerular chronic kidney disease and those with nonglomerular chronic kidney disease. Am J Epidemiol 174:604–612. 10.1093/aje/kwr121 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Schneider MF, Muñoz A, Ku E, Warady BA et al. (2021) Estimation of Albumin-Creatinine Ratio From Protein-Creatinine Ratio in Urine of Children and Adolescents With CKD. Am J Kidney Dis 77:824–827 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Schauer PR, Bhatt DL, Kirwan JP, Wolski K et al. (2017) Bariatric Surgery versus Intensive Medical Therapy for Diabetes — 5-Year Outcomes. N Engl J Med 376:641–651. 10.1056/nejmoa1600869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Navaneethan SD, Yehnert H, Moustarah F, Schreiber MJ et al. (2009) Weight loss interventions in chronic kidney disease: A systematic review and meta-analysis. Clin J Am Soc Nephrol 4:1565–1574. 10.2215/CJN.02250409 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Saland JM, Pierce CB, Mitsnefes MM, Flynn JT et al. (2010) Dyslipidemia in children with chronic kidney disease. Kidney Int 78:1154–1163. 10.1038/ki.2010.311 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sgambat K, Clauss S, Lei KY, Song J et al. (2018) Effects of obesity and metabolic syndrome on cardiovascular outcomes in pediatric kidney transplant recipients: a longitudinal study. Pediatr Nephrol 33:1419–1428. 10.1007/s00467-017-3860-8 [DOI] [PubMed] [Google Scholar]
- 31.Zhang L, Wang Y, Xiong L, Luo Y et al. (2019) Exercise therapy improves eGFR, and reduces blood pressure and BMI in non-dialysis CKD patients: Evidence from a meta-analysis. BMC Nephrol 20:398. 10.1186/s12882-019-1586-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.O’Leary VB, Marchetti CM, Krishnan RK, Stetzer BP et al. (2006) Exercise-induced reversal of insulin resistance in obese elderly is associated with reduced visceral fat. J Appl Physiol 100:1584–1589. 10.1152/japplphysiol.01336.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Alp Ikizler T, Robinson-Cohen C, Ellis C, Headley SAE et al. (2018) Metabolic effects of diet and exercise in patients with moderate to severe CKD: A randomized clinical trial. J Am Soc Nephrol 29:250–259. 10.1681/ASN.2017010020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Clark SL, Denburg MR, Furth SL (2016) Physical activity and screen time in adolescents in the chronic kidney disease in children (CKiD) cohort. Pediatr Nephrol 31:801–808. 10.1007/s00467-015-3287-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Shah PP, Brady TM, Meyers KEC, O’Shaughnessy MM et al. (2021) Association of obesity with cardiovascular risk factors and kidney disease outcomes in primary proteinuric glomerulopathies. Nephron 145:245–255. 10.1159/000513869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Staples AO, Greenbaum LA, Smith JM, Gipson DS et al. (2010) Association between clinical risk factors and progression of chronic kidney disease in children. Clin J Am Soc Nephrol 5:2172–2179. 10.2215/CJN.07851109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Furth SL, Pierce C, Hui WF, White CA et al. (2018) Estimating Time to ESRD in Children With CKD. Am J Kidney Dis 71:783–792. 10.1053/j.ajkd.2017.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Babayev R, Whaley-Connell A, Kshirsagar A, Klemmer P et al. (2013) Association of race and body mass index with ESRD and mortality in CKD stages 3–4: Results from the kidney early evaluation program (KEEP). Am J Kidney Dis 61:404–412. 10.1053/j.ajkd.2012.11.038 [DOI] [PubMed] [Google Scholar]
- 39.Lin T-Y, Peng C-H, Hung S-C, Tarng D-C (2018) Body composition is associated with clinical outcomes in patients with non–dial-ysis-dependent chronic kidney disease. Kidney Int 93:733–740. 10.1016/j.kint.2017.08.025 [DOI] [PubMed] [Google Scholar]
- 40.Ahmadi SF, Zahmatkesh G, Ahmadi E, Streja E et al. (2015) Association of Body Mass Index with Clinical Outcomes in Non-Dialysis-Dependent Chronic Kidney Disease: A Systematic Review and Meta-Analysis. CardioRenal Med 6:37–49. 10.1159/000437277 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.De Nicola L, Minutolo R, Chiodini P, Borrelli S et al. (2012) The effect of increasing age on the prognosis of non-dialysis patients with chronic kidney disease receiving stable nephrology care. Kidney Int 82:482–488. 10.1038/ki.2012.174 [DOI] [PubMed] [Google Scholar]
- 42.Yun HR, Kim H, Park JT, Chang TI et al. (2018) Obesity, Metabolic Abnormality, and Progression of CKD. Am J Kidney Dis 72:400–410. 10.1053/j.ajkd.2018.02.362 [DOI] [PubMed] [Google Scholar]
- 43.Lin T-Y, Lim P-S, Hung S-C (2018) Impact of Misclassification of Obesity by Body Mass Index on Mortality in Patients With CKD. Kidney Int Rep 3:447–455. 10.1016/j.ekir.2017.12.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Sharma D, Hawkins M, Abramowitz MK (2014) Association of Sarcopenia with eGFR and Misclassification of Obesity in Adults with CKD in the United States. Clin J Am Soc Nephrol 9:2079–2088. 10.2215/CJN.02140214 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Lassalle M, Fezeu LK, Couchoud C, Hannedouche T et al. (2017) Obesity and access to kidney transplantation in patients starting dialysis: A prospective cohort study. PLoS One 12:e0176616. 10.1371/JOURNAL.PONE.0176616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Beddhu S, Pappas LM, Ramkumar N, Samore M (2003) Effects of body size and body composition on survival in hemodialysis patients. J Am Soc Nephrol 14:2366–2372 [DOI] [PubMed] [Google Scholar]
- 47.Gracia-Iguacel C, Qureshi AR, Avesani CM, Heimburger O et al. (2013) Subclinical versus overt obesity in dialysis patients: more than meets the eye. Nephrol Dial Transplant 28:iv175–iv181 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.