PEDIATRICS Vol. 109 No. 5 May 2002, pp. 909-913
Adverse Clinical Outcomes Associated With Short Stature at Dialysis Initiation: A Report of the North American Pediatric Renal Transplant Cooperative Study
,¶,#
* Department of Pediatrics, Johns Hopkins Medical Institutions, Baltimore, Maryland
Welch Center for Prevention, Epidemiology and Clinical Research, Johns Hopkins Medical Institutions, Baltimore, Maryland
¶ Department of Epidemiology, Johns Hopkins Medical Institutions, Baltimore, Maryland
# Department of Medicine, Johns Hopkins Medical Institutions, Baltimore, Maryland
EMMES Corporation, Potomac, Maryland
|| Department of Pediatrics, State University of New York at Stony Brook, New York
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective. We examined whether height less than the 1% for age (z score <-2.5) at dialysis initiation predicts adverse clinical outcomes for children with kidney failure.
Design. National cohort study of children initiating dialysis, followed for a minimum of 1 month to a maximum of 8 years.
Setting. Voluntary consortium of pediatric nephrology centers across the United States and Canada in the North American Pediatric Renal Transplant Cooperative Study.
Patients. Two thousand three hundred six patients 
21 years old initiated on dialysis between 1992 and 2000.
Outcome Measures. School attendance, transplant wait listing, hospitalization rates, and survival.
Results. Although there were no differences in transplant wait listings, school-aged children with short stature were less likely to be attending school full-time than were their counterparts with more normal height, even if medically capable. Short-stature patients have significantly more hospital days per month of dialysis follow-up than those patients with better growth (mean: 1.92 vs 1.58; median: 0.73 vs 0.44 hospital days per month of follow-up). Cox proportional hazards regression analyses show that children with height <1% for age have a twofold higher risk of death than those with more normal height, even after controlling for patient age, race, gender, cause of end-stage renal disease, wait list status, and dialysis modality.
Conclusions. Poor growth during chronic renal insufficiency leading to short stature at dialysis initiation is a marker for a more complicated clinical course for children with kidney failure. Aggressive nutritional support and early referral to a nephrologist to optimize growth may improve long-term outcomes for children with chronic kidney disease.
Key Words: growth • kidney disease • morbidity
Abbreviations: NAPRTCS, North American Pediatric Renal Transplant Cooperative Study • CRI, chronic renal insufficiency • ESRD, end-stage renal disease • SD, standard deviation • RH, relative hazard • CI, confidence interval • SDS, standard deviation score
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Growth failure is a significant problem for children with chronic kidney disease. Reports from the US Renal Data System1 and the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS)2,3 demonstrate substantial height deficits not only for children on dialysis but also for those with chronic renal failure and posttransplant. Growth failure among these patients is likely multifactorial, relating to the age at which chronic renal insufficiency (CRI) develops, the degree of acidosis, insufficient caloric intake, renal osteodystrophy, and abnormalities of the growth hormone/insulin-like growth factor axis in renal failure.4, 5 It remains unclear whether growth problems in children with chronic kidney disease are entirely remediable through medical management. However, it has been shown that aggressive nutritional intervention—in combination with management of anemia, acidosis, and bone disease and use of recombinant human growth hormone—can substantially improve growth in patients with CRI.5 Recent reports suggest that short stature and poor growth in children maintained on dialysis for end-stage renal disease (ESRD) are associated with an increased risk of death.6 Our goal in this study was to determine whether short stature at dialysis initiation is an independent marker for poor clinical outcome. We presumed that short stature at dialysis initiation reflects poor growth during progressive renal insufficiency. Documentation of the ill effects of poor growth before development of ESRD would highlight the importance of identifying patients with CRI for medical intervention and possibly early referral to a nephrologist.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study Design, Population, and Data Sources
We conducted a cohort study of children enrolled in the NAPRTCS who began dialysis between January 1, 1992, and January 17, 2001. Since January 1, 1992, data has been voluntarily collected on children who have had dialysis at over 150 centers in North America.2,3 Participating NAPRTCS centers are listed in the most recent NAPRTCS Annual Report.2 Patients enrolled in the dialysis arm of the NAPRTCS study are
21 years of age at their index initiation of dialysis. Demographic data obtained at registration includes patient age, race, gender, and primary renal diagnosis. Follow-up data are obtained 30 days after initiation of dialysis and every 6 months thereafter until patients receive transplants, they switch dialysis modalities (eg, from peritoneal to hemodialysis), their renal function improves, or they die. A standardized height assessment and school status report is included at the follow-up report 30 days after dialysis initiation. If the patient has not yet completed a high school education and is of school age, the health care provider is asked to code whether the child attends school full-time or part-time, is home schooled, or if not attending school, is medically capable or incapable of attending school. Providers are also asked whether a patient is on the transplant waiting list, in the process of a transplant workup, or, if not on the list, not listed for a medical reason or patient/family preference. Patients without a standardized height assessment were excluded from the study. Patients were followed after dialysis initiation for a minimum of 1 month to a maximum of 8 years.
Analysis
Cross-sectionally, we examined whether short stature at initiation of dialysis is associated with lower rates of school attendance or differences in transplant wait list status. Longitudinally, we examined whether short stature at initiation of dialysis predicts death, or increased risk of hospitalization using survival analysis (comparing groups by height, without accounting for other variables) and Cox proportional hazards analysis. In proportional hazards analysis, the relative hazard gives an estimate of the overall difference between the survival curves of patients with short stature and those with normal height after controlling for other potential confounding factors.
Short stature was categorized as height <-2.5 standard deviations (SD) below age norms for height. We compared hospitalizations and death rates for patients with height <-2.5 SD with those for patients >-2.5 SD below norms for age. The height cutoff of <-2.5 SD below age norms includes patients who are less than rather than 1% in height for age. We also included height z score as a continuous (rather than categorical) variable in a separate model.
School attendance was examined by comparing the proportions of patients with height z < or >-2.5 SD who were attending school full-time, part-time, or home schooling by 
2. We compared crude death rates to examine mortality according to short stature. Cox proportional hazard models were used to assess whether short stature was an independent risk factor for death, while controlling for patient age, gender, race, transplantation waiting list status, dialysis modality at initiation, and underlying cause of renal disease. Treatment era (1992–1996 vs 1996–2000) was also included in a subsequent model.
Underlying cause of renal disease was categorized broadly into 3 groups as follows: 1) Structural: aplasia, hypoplasia, dysplasia, and obstructive uropathy; 2) Focal and segmental sclerosis, membranous, proliferative, idiopathic, crescentic or chronic glomerulonephritis, pyelonephritis, interstitial nephritis, hemolytic uremic syndrome, familial nephritis; and 3) Other: polycystic kidney disease, congenital nephrotic syndrome, cystinosis, and systemic immunologic disease.
Morbidity for patients with short stature was scored by counting the number of hospitalization days, scaling by the months of follow-up and ranking patients based on these. Multiple regression analysis of the ranked morbidity score was performed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the NAPRTCS cohort, 2424 patients initiated dialysis as their first ESRD modality between 1992 and January 2001. Of these patients, 118 did not have a height assessment at their 30-day report. The remaining 2306 subjects were eligible for our study. Demographic characteristics of the study group are presented in Table 1, stratified by height 30 days after initiation of dialysis. The patients with the greatest height deficit tended to be younger and were more often male. In addition, patients with diagnoses of structural abnormalities, (including aplasia, hypoplasia, dysplasia, and obstructive uropathy) were more likely to be short. They were also more frequently initiated on peritoneal dialysis. Only 15% of the short-stature group was black compared with 25% of those with normal height.
|
Table 2 demonstrates the results of the cross-sectional analysis. At dialysis initiation patients with height <-2.5 SD below normal for age were less likely to be attending school full-time and were more likely to be medically capable but not attending school. There was no difference in transplant waiting list status by height at dialysis initiation.
|
Table 3 presents the risk of death for patients with short stature at dialysis initiation. The crude death rate for individuals <-2.5 SD below age norms was 14.7% (84/573) compared with 6.1% (105/1733) in the patients with height closer to the normal range. Because young patient age has been shown to be an important risk factor for mortality, we also performed an analysis stratified by age group. Poor growth was a risk factor for death in both the young (age <1 year) and older age group (Fig 1). The risk of death associated with short stature persisted in an analysis of mortality after adjustment for age, race, modality, diagnosis, and wait list status (Table 3). Children initiating dialysis with height z score <-2.5 were more than twice as likely to die during follow-up than patients with height in the normal range (adjusted relative hazard [RH]: 2.07; 95% confidence interval [CI]: 1.53–2.79). This result was virtually unchanged when treatment era was considered, and treatment era was not significant predictor of survival. A similar result was observed when stature is used as a continuous variable with a relative hazard of 0.88 (P = .002) per SD increase in stature. Age <1 year was also a strong independent risk factor for death. Infants initiating dialysis at <1-year-old were >3 times as likely to die as were older children (adjusted RH 3.07; 95% CI: 2.28–4.12). This was true even if other patient characteristics—height, race, gender, and grouped cause of ESRD—were identical. In addition, compared with children with structural causes of ESRD, children with "other" causes of ESRD had an increased risk of death (adjusted RH 1.91; 95% CI: 1.34–2.71). This category is heterogeneous and includes systemic immunologic disease, polycystic kidney disease, congenital nephrotic syndrome, cystinosis, and Denys Drash. Each cause within the category is relatively small, allowing for only imprecise estimates of the relative risk associated with each one of these diagnoses.
|
|
Overall, dialysis patients were hospitalized a mean of 1.66 days per month of follow-up (median of 0.50 days). Patients with short stature (z <-2.5 SD) had significantly more hospital days per month than patients with height z > -2.5 SD (median 0.73 compared with 0.44, P < .001; mean 1.92 vs 1.58; Table 4). In an analysis of ranked outcomes, age (P < .001), wait list status (P < .001), race (P = .05), and short stature (P = .02), but not gender, diagnosis, or dialysis type were significant predictors of hospitalization time. Adjusted analysis of the ranked morbidity score showed that after controlling for age, race, gender, cause of ESRD, wait list status and dialysis modality, patients with short stature spent an average of 0.22 more days per month in the hospital than their counter parts with more normal height (Table 4).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Poor growth is a well-recognized complication of ESRD in children.1,6, 7 Poor nutritional status is a well-known risk factor for infectious complications and mortality in adults with ESRD.8–10 Echoing the findings in adults, a recent study by Wong et al6 demonstrated that among pediatric patients with ESRD, each standard deviation score (SDS) decrease in height among patients maintained on dialysis or posttransplant was associated with a 14% (95% CI: 9%–34%) increase in death. Our study demonstrates that poor growth preceding the onset of ESRD, during progressive renal insufficiency, is also an important risk factor for adverse clinical outcomes. Heights on the patients in this study were determined shortly after dialysis initiation, and patients were followed longitudinally for risk of hospitalization and mortality. In cross-sectional analysis, we found that patients who were younger, male, and had obstructive uropathy or dysplasia as a cause of ESRD were more likely to be short at dialysis initiation. This is likely related to early age of onset of CRI in this group of patients. Our findings point to the importance of maximizing growth as early as CRI is noted, before initiation of dialysis, as these patients with short stature are at increased risk of a more complicated clinical course. In addition to increasing risk of hospitalization and death, significant height deficits before the initiation of dialysis are associated with poor school attendance, an important marker of functional status for children.11,12
The pathogenesis of growth retardation in children with chronic renal failure is multifactorial, related to nutritional deficits, acidosis, anemia, bone disease, uremia, and relative resistance to growth hormone.5 Our study cannot specify which combination of these factors mediates the increased risk of adverse outcomes seen in this population. Several possibilities exist. Inadequate nutrition during CRI could be associated with persistent nutritional deficits while on dialysis. Alternatively nutritional deficiencies during CRI could make patients more vulnerable later, even if nutrition improves. Complex medical regimens including bicarbonate therapy, iron, epoetin 
, salt and water supplementation, and growth hormone are frequently necessary to ensure optimal growth in pediatric CRI patients.4,5,13, 14 Short stature may be a marker for those patients who are not identified for early, aggressive intervention. It may identify those patients who face barriers to pre-ESRD care, receive suboptimal medical management once care is accessed, or who are unable to adhere to the complex medical regimen necessary for these patients to thrive. Finally, biological factors, which prevent optimal growth even in those patients with optimal medical management and adherence, may contribute to poorer outcomes in this patient group.
This study has several limitations that deserve mention. The NAPRTCS is a voluntary consortium of pediatric nephrology centers and does not collect data on every pediatric patient with ESRD in North America. However, pediatric patients in NAPRTCS are thought to constitute >80% of all pediatric ESRD patients in the United States, and potential differences between enrollees in NAPRTCS and other pediatric patients with ESRD should not affect the internal validity of our results. In addition, as we only analyzed height at initiation of dialysis, we cannot assess how intervening treatment decisions and management affect ultimate patient outcomes.
Despite these limitations, the strength of this analysis lies in its national scope and prospective data collection of the NAPRTCS. In addition, the large number of children initiating dialysis in this multicenter cohort allows for multivariate analysis of risk. This analysis demonstrates that poor growth before dialysis initiation predicts adverse clinical outcomes in ESRD. Although this study does not delineate the exact factors mediating adverse outcomes for children with short stature, it points to the crucial importance of aggressive pre-ESRD care in maximizing long-term health. Research addressing barriers to patients accessing care before developing ESRD, the optimization of nutrition, and the management of anemia, acidosis, bone disease, and growth during progressive CRI may decrease morbidity and mortality for children with kidney failure. In addition, interventions designed to improve patient adherence to the complex treatment regimens necessary to optimize growth are of utmost importance to maximize health outcomes for children with kidney failure.
|
ACKNOWLEDGMENTS |
|---|
Dr Furth is supported by grant K08 DK02586–01A1 and Dr Powe by grant K24 DK02643 from the National Institute of Diabetes and Digestive and Kidney Disease, Bethesda, Maryland. Portions of this work were funded by the Genentech Foundation for Growth and Development. NAPRTCS is a voluntary collaborative effort comprising more than 150 pediatric renal disease treatment centers in the United States, Canada, Mexico, and Costa Rica. It is supported by major, unrestricted educational grants from Novartis, Amgen, and Genentech.
|
FOOTNOTES |
|---|
Received for publication Aug 24, 2002; Accepted Jan 15, 2002.
Reprint requests to (S.L.F.) NAPRTCS Operations Manager, 19 Bradhurst Ave, Box 10, Hawthorne, NY 10532-2140. E-mail: sfurth{at}jhmi.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
- Turenne MN, Port FK, Strawderman RL, et al. Growth rates in pediatric dialysis patients and renal transplant recipients. Am J Kidney Dis.1997; 30 :193 –203[Medline]
- Stablien DM, Ho M. North American Pediatric Renal Transplant Cooperative Study 2001 Annual Report. Potomac, MD: EMMES Corporation;2001 . Available at: http://spitfire.emmes.com/study/ped/annlrept/index.htm. Accessed March 26, 2002
- Seikaly M, Ho PL, Emmett L, Tejani A. The 12th Annual Report of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS): renal transplantation from 1987 through 1998. Pediatr Transplant.2001; 5 :215 –231[CrossRef][Medline]
- Warady BA, Alexander, Watkins S, Kohaunt E, Harmon WE. Optimal care of the pediatric end-stage renal disease patient on dialysis. Am J Kidney Dis.1999; 33 :567 –583[Medline]
- Furth SL, Alexander D, Hwang W, Neu AM, Powe NR, Fivush BA. Does poor growth predict morbidity and mortality in children with chronic renal failure and on dialysis? Semin Nephrol.2001; 21 :463 –469[CrossRef][Medline]
- Wong CS, Gipson DS, Gillen DL, et al. Anthropometric measures and risk of death in children with end-stage renal disease. Am J Kidney Dis.2000; 36 :811 –819[Medline]
- Abitbol C, Foreman JW, Strife CF, McEnery PT. Quantitation of growth deficits in children with renal diseases. Semin Nephrol.1989; 9 :31 –36[Medline]
- Held PJ, Port FK, Gaylin DS, et al. Comorbid conditions and correlation’s with mortality risk among 3399 incident hemodialysis patients. Am J Kidney Dis.1992; 20(suppl 2) :32 –38[Medline]
- Churchill DN, Taylor DW, Cook RJ, et al. Canadian Hemodialysis Morbidity Study. Am J Kidney Dis.1992; 19 :214 –234[Medline]
- Bergstrom J. Nutrition and mortality in hemodialysis [editorial]. J Am Soc Nephrol.1995; 6 :1329 –1341[Abstract]
-
Henning P, Tomlinson L, Rigden SP, Haycock GB, Chantler C. Long term outcome of treatment of end stage renal disease.
Arch Dis Child.1988; 63
:35
–40
[Abstract/Free Full Text] - Morthon MJS, Reynolds JM, Garralda ME, et al. Psychiatric adjustment in end stage renal disease: a follow-up study of former pediatric patients. J Psychosom Res.1994; 38 :293 –303[CrossRef][Medline]
-
Haffner D, Schaefer F, Nissel R, et al. Effect of growth hormone treatment on the adult height of children with chronic renal failure.
N Engl J Med.2000; 343
:923
–930
[Abstract/Free Full Text] - Rees L, Ward G, Ridgen SP. Growth over 10 years following a 1 year trial of growth hormone therapy. Pediatr Nephrol.2000; 14 :309 –314[CrossRef][Medline]
PEDIATRICS (ISSN 1098-4275). ©2002 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  B. J Foster and M. B Leonard Measuring nutritional status in children with chronic kidney disease Am. J. Clinical Nutrition, October 1, 2004; 80(4): 801 - 814. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Goldstein, C. M. Smith, and H. Currier Noninvasive Interventions to Decrease Hospitalization and Associated Costs for Pediatric Patients Receiving Hemodialysis J. Am. Soc. Nephrol., August 1, 2003; 14(8): 2127 - 2131. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
