Comparison of growth status of patients with cystic fibrosis between the United States and Canada 1 3

Transcription

1 Comparison of growth status of patients with cystic fibrosis between the United States and Canada 1 3 Hui-Chuan Lai, Mary Corey, Stacey FitzSimmons, Michael R Kosorok, and Philip M Farrell ABSTRACT Background: Differences in growth status of patients with cystic fibrosis (CF) between the United States and Canada were reported in the 1980s based on analysis of data from 2 regional CF centers. Objective: We evaluated the current growth status of the entire CF population in the United States and Canada in view of recent advances in the treatment of CF. Design: Growth data from the CF Patient Registries were analyzed. Results: Mean height and weight were at approximately the 30th percentile for children with CF in the United States. Mean height and weight were 4 5 percentiles higher in children with CF in Canada than in those in the United States (P < 0.01), but percentages of ideal weight (104%) were similar in both populations. In adults with CF, mean height was similar at the 37th percentile; however, weight (26th compared with the 21st percentiles) and percentage of ideal weight (93% compared with 90%) were significantly higher in Canada than in the United States. Differences related to sex and age were similar in both countries for all indexes, which showed a high prevalence of underweight in infants and in older patients, but little sex discrepancy. Conclusion: We observed substantially smaller differences in the growth indexes of CF patients between the United States and Canada compared with results from the 1980s. These findings reflect significant improvements in the nutritional status of US patients in recent years. However, caution is required in the direct comparison of mean percentiles from reports using different growth standards because there are systematic differences in growth standards, which affect, in particular, the comparison of growth in males and females. Am J Clin Nutr 1999;69: KEY WORDS Cystic fibrosis, growth, height percentile, weight percentile, nutritional status, malnutrition, United States, Canada, humans INTRODUCTION Cystic fibrosis (CF) is the most common severe autosomal, recessive genetic disorder with estimated incidences of 1 in 2900 white newborns in the United States (1) and 1 in 2500 in Canada (2). Growth failure and malnutrition, resulting from multiple, interrelated factors, are common and well documented to be unfavorable prognostic factors in CF (3 5). Assessment of growth and nutritional status in large CF populations can be useful in the development and evaluation of intervention strategies (6). Differences in the survival and growth status of the CF population between the United States and Canada have been documented. Corey et al (7) reported substantially better growth of CF patients in Toronto (eg, mean height-for-age and weight-forage of 42nd and 43rd percentiles for males, respectively) than in Boston (eg, mean height-for-age and weight-for-age of 33rd and 35th percentiles for males, respectively) using data obtained in In the same study (7), survival curves after 10 y of age were also better in Toronto than in Boston. It was hypothesized that differences in nutrition therapy, ie, no restriction of dietary fat in Toronto, could explain the differences in growth and survival between these 2 CF populations (7). An updated comparison of the growth status of CF patients between the United States and Canada is important because previous results obtained from 2 local CF centers may not represent the entire CF populations in these 2 countries. In addition, significant advances in the treatment of CF have occurred over the past 2 decades, resulting in concurrent improvement in the survival and clinical status of CF patients. Our recent analyses of the 1993 US Cystic Fibrosis Patient Registry showed that the prevalence of malnutrition is very high in children with newly diagnosed CF (44% below the 5th percentile for height or weight); however, only 24% of children with CF who had received ongoing medical treatment were below the 5th percentile for height or weight (8). The 1996 report from the US Cystic Fibrosis Foundation also documented an increase in the median survival age from 26.9 to 31.3 y and a 20% decrease in 1 From the Departments of Pediatrics and Biostatistics, University of Wisconsin School of Medicine, Madison; the Division of Gastroenterology and Nutrition, Departments of Public Health Sciences and Pediatrics, University of Toronto, The Hospital for Sick Children; and the Cystic Fibrosis Foundation, Bethesda, MD. 2 Supported by NIH grants R01 DK34108 and a Postdoctoral Research Fellowship from the National Cystic Fibrosis Foundation. 3 Address reprint requests to H-C Lai, Departments of Pediatrics and Biostatistics, University of Wisconsin Madison, K6/428 Clinical Sciences Center, Madison, WI lai@biostat.wisc.edu. Received April 1, Accepted for publication August 7, Am J Clin Nutr 1999;69: Printed in USA American Society for Clinical Nutrition 531

2 532 LAI ET AL the prevalence of malnutrition from 1986 to 1996 (9). Furthermore, the different approaches to nutrition therapy documented in the 1970s between the United States and Canada (7, 10, 11) appear to no longer exist now that nutrition intervention techniques are well developed (12, 13). In particular, a fat-enriched diet and improved pancreatic enzyme replacement therapy have become the standard nutrition practices in US CF centers in the current decade (14). In view of these advances, we compared the growth status and prevalence of malnutrition between US and Canadian CF patients. SUBJECTS AND METHODS Cystic Fibrosis Patient Registry data The Cystic Fibrosis Patient Registry operated by the US Cystic Fibrosis Foundation and the Canadian Patient Data Registry operated by the Canadian Cystic Fibrosis Foundation document the diagnosis and annual follow-up of all known CF patients in the United States and Canada. Basic demographic data, diagnostic information, and selected clinical characteristics are collected annually, as described in detail elsewhere (2, 15). Demographic (race, sex, and age), growth (height and weight), and select clinical characteristics (age at diagnosis, history of meconium ileus, genotype, and use of pancreatic enzyme supplementation to reflect pancreatic functional status) of patients with complete growth measurements for the years 1992, 1993, and 1994 were obtained from both registries. Approximately 1% of US records and 8% of Canadian records had missing height, weight, or both measurements for a given year and these records were not included in our study. Computation of growth indexes The following growth indexes were calculated to reflect nutritional status: 1) height and weight percentiles, 2) height and weight z scores, and 3) percentages of ideal weight. For pediatric CF patients (<18 y of age), height and weight percentiles and z scores were calculated by using the EPI INFO program (16), which uses normalized growth reference curves developed by the National Center for Health Statistics (NCHS) (17 19). For adult CF patients ( 18 y of age), height and weight percentiles and z scores were calculated by using reference values from the second National Health and Nutrition Examination Survey (NHANES II) (20). Both NCHS standards for children and NHANES II references for adults are used as the reference populations in the recommended dietary allowances (21). Percentage of ideal weight, as recommended by the US Cystic Fibrosis Foundation Nutrition Consensus Report (12), was calculated by using our previously developed SAS program, which uses reference values (ie, mean weights and their corresponding SDs) derived from NCHS (17) and NHANES II (20) standards for this computation (8). Determination of nutritional status Frequency distributions of height and weight percentiles were categorized into 6 percentile channels (<5th, 5 10th, 10 25th, 25 50th, 50 75th, and >75th) to classify nutritional status. Percentage of ideal weight distribution (12) was categorized as >110% (overweight), % (normal weight), 85 90% (underweight), 80 85% (mild malnutrition), 75 80% (moderate malnutrition), and <75% (severe malnutrition) to determine the degree of malnutrition. In addition to the above classifications for nutritional status, mean values for each growth index in specific age groups were evaluated in relation to sex and age. Predetermined cutoff points of growth indexes were used to define different indicators of malnutrition: height and weight below the 5th percentile indicated stunting and wasting, respectively, and weight <85% of ideal indicated nutritional failure. Data and statistical analyses There were 1379 US records (2.6%) and 108 Canadian records (1.3%) with height or weight values that deviated >4 SDs from the reference mean of the NCHS reference population. These height and weight values were thought to be outliers resulting from measurement or recording errors; eg, it is highly unlikely for a 10-y-old child to be 50 cm tall or for an infant to weigh 80 kg. In addition, these height and weight values could not be computed accurately into age- and sex-specific percentiles by the EPI INFO program (16); therefore, these values were excluded from growth analyses. SAS (version 6.12, 1996; SAS Institute, Inc, Cary NC) and S- PLUS statistical software (StatSci, Division of MathSoft, Inc, Seattle) were used to compile descriptive tables, summary statistics, and graphic analyses. All statistical analyses were performed separately for each sex or stratified by sex because of the well-documented sex difference in survival and clinical course between male and female CF patients (7, 22 24). Comparisons between categoric outcomes were assessed by using chi-square contingency tables (25) in SAS. The method of generalized estimating equations (GEE) (26) was used to compare group means, adjusted for repeated measures and stratified by age when appropriate. Group differences were assessed by t tests of the relevant regression coefficients. The identity link was used for continuous outcomes (eg, growth percentiles and z scores) and the logit link was used for dichotomous outcomes (eg, indicators for falling below the 5th percentile for height and weight). An SAS macro implementation of GEE by Karim and Zeger (27) was used for GEE analyses. Last, P values for comparisons of medians between groups were obtained from the median test of Su and Wei (28) by using S-PLUS. The following null hypothesis was tested: quantiles from the 2 samples are equal but corresponding distribution functions may not be equal. RESULTS Patient characteristics Demographic and diagnostic characteristics of the study populations are summarized in Table 1. Most CF patients were white; however, the percentage of African American patients was significantly higher and that of patients of other races lower in the United States than in Canada in both all patients and in patients with a new diagnosis of CF. The mean age for all patients was slightly yet significantly lower in the United States than in Canada and the percentage of pediatric patients was higher in the United States (70%) than in Canada (67%). More than 75% of the patients with a new diagnosis of CF were <4 y of age, and the age distribution was similar between the United States and Canada in this subpopulation. Comparison of the diagnostic characteristics between the 2 countries showed that age at diagnosis was similar. However, proportionately fewer patients with meconium ileus and with pancreatic enzyme supplementation, but more patients with non

3 GROWTH IN PATIENTS WITH CYSTIC FIBROSIS 533 TABLE 1 Demographic and diagnostic characteristics of cystic fibrosis (CF) patients reported to the CF Patient Registries of the United States (US) and Canada 1 All patients New cases 2 Variable US Canada US Canada Demographic characteristics Number of patients Male sex (%) Race (%) 4,5 White African American Other Age (y) x + SD 14.0 ± ± ± ± 8.4 Median Age distribution (%) 4 Infants (< 1 y) Children 1 3 y y y Adolescents y y Adults y y y Age at diagnosis (y) x + SD 2.9 ± ± ± ± 8.4 Median Meconium ileus (%) 4, Patients with genotype data (%) 4, Genotype (%) 4,5 F508/ F F508/other Other/other Patients with pancreatic enzyme supplementation data (%) 4, Use of enzymes (%) 4,5 Yes No Values for the time-dependent variables age and age distribution were computed by using all records, which may include multiple observations per patient; values for all other variables were computed by using the most recent observation of each patient. 2 Values were computed by using data obtained during the year of diagnosis and included only one observation per patient. 3 About 70% of patients had growth measurements for all 3 y; the remaining patients had 2 ( 16% of patients) and 1 ( 14% of patients) measurements. The numbers of observations for 1992, 1993, and 1994 were , , and for the US and 2687, 2743, and 2811 for Canada. The numbers of new cases in 1992, 1993, and 1994 were 959, 884, and 891 for the US and 105, 110, and 119 for Canada. 4,5 Significant difference among quintiles, P < 0.05: 4 all patients, 5 new cases. 6 Significantly different from US (all patients), P < F508 mutations, were observed in Canada than in the United States in both all patients and in patients with a new diagnosis of CF. The relatively large proportion of Canadian patients with missing enzyme data were accounted for by a single large center and was therefore not likely to explain differences in the percentage using supplements between the 2 countries. Overall growth status The overall growth status of CF patients in the United States and Canada is compared in Figure 1. Frequency distributions of height and weight percentiles were skewed to the left compared with the expected distribution of the NCHS reference population, indicating subnormal growth of CF patients. Mean and median height and weight percentiles were consistently lower in the United States than in Canada. Although the differences were small (2 4 percentiles for means and 4 6 percentiles for medians), they were significant. In addition, the percentage of patients with height below the 5th percentile and weight below the 5th percentile was also 4 6% higher (P < 0.01, GEE) in the United States than in Canada. Differences in the percentage of ideal weight were much less pronounced; only the mean value for males was significantly lower (P = 0.01, GEE) in the United States than in Canada. Growth status in relation to sex and age Mean height z scores, weight z scores, and percentage of ideal weight are shown in Figure 2. The percentage of patients with malnutrition as indicated by height below the 5th percentile, weight below the 5th percentile, and weight <85% of ideal in

4 534 LAI ET AL FIGURE 1. Comparison of growth status between US (, n = males and females) and Canadian (Can;, n = 4315 males and 3816 females) cystic fibrosis (CF) patients as indicated by frequency distributions of height percentiles, weight percentiles, and percentages of ideal weight by using National Center for Health Statistics standards for children (16) and second National Health and Nutrition Examination Survey references for adults (19). * Significantly different from the United States (US), P < 0.05 (generalized estimating equations and median tests were used for means and medians, respectively). Data are from the CF Patient Registries of the United States and Canada. relation to sex and age is shown in Figure 3. Specifically, growth was assessed for 4 distinct age periods: infancy to 2 y of age, early childhood, preadolescence to adolescence, and adulthood. Infancy to 2 y In infants with CF, mean height and weight z scores were below 1 SD (corresponding approximately to the 15th percentile) and 30 40% of patients were below the 5th percentile for height and weight for both sexes in both countries. Height and weight z scores appeared to improve from infancy to 2 y. In 2-y-old children with CF, mean height and weight z scores were 44 79% higher (range: 0.2 to 0.6) than those of infants with CF and <20% of the 2-y-olds were below the 5th percentile for height and weight. GEE analyses indicated that males aged 0 2 y in Canada had significantly higher weight z scores than males aged 0 2 y in the United States and that for males aged 0 2 y in Canada, more were of ideal weight, fewer had heights less than the 5th percentile, and fewer were <85% of ideal weight compared with their US counterparts. Females in Canada had significantly higher weight z scores than females in the United States and the differences appeared to be smaller than those in males. In addition, no differences in the prevalence of malnutrition ie, the percentage of patients with height below the 5th percentile, weight below the 5th percentile, or <85% of ideal weight were observed between countries in females aged 0 2 y. Early childhood For CF patients aged 3 10 y, males in Canada had consistently and significantly higher height z scores (Canada: from 0.3 to 0.7 ; United States: from 0.5 to 0.8) and weight z scores (Canada: from 0.2 to 0.5; United States: from 0.4 to 0.7); fewer males in Canada than in the United States had heights and weights below the 5th percentile. Mean percentages of ideal weight were normal ( % of ideal weight) and <5% of patients were <85% of their ideal weight. No differences in percentages of ideal weight were noted between the United States and Canada in males aged 3 10 y. Similar trends in the relation between age and height and weight z scores were observed in females aged 3 to 7 8 y; however, percentages of ideal weight were significantly higher in the United States than in Canada in this age range. Preadolescence to adolescence Dramatic changes in growth indexes and prevalence of malnutrition were observed in CF patients aged 7 8 y through 18 y. Males with CF showed consistent and rapid declines in mean height and weight z scores from 9 10 y to y, with some improvement in height but not in weight at y. Percentages of ideal weight remained normal (>100% of ideal weight) until age y. Similar declines in height and weight z scores were observed in female CF patients from 7 8 y to y, with improvement in height and stabilization in weight from 12 to 18 y. Weight z scores appeared to decline at earlier ages than did height z scores. Percentages of ideal weight declined from 105% at age 12 y to 93% at age 18 y. For this age group, US patients again had significantly lower mean height and weight z scores than the Canadians, but percentages of ideal weight were similar between the 2 countries.

5 GROWTH IN PATIENTS WITH CYSTIC FIBROSIS 535 FIGURE 2. Comparison of mean (± SD) height z scores, weight z scores, and percentages of ideal weight by sex and age between cystic fibrosis (CF) patients in the United States and Canada. The corresponding number of observations are indicated for each of the 4 age subgroups. Significance was assessed with generalized estimating equations. Data are from the CF Patient Registries of the United States and Canada. Parallel changes in the prevalence of malnutrition were observed during this age period. In particular, 4 7 times as many 18- than 10-y-old males and 2 5 times as many 18- than 10-y-old females had weights less than the 5th percentile and were <85% of their ideal weight. In 18-y-old patients, 30 40% were below the 5th percentile for weight and were <85% of ideal weight. Adulthood In adults aged y, height appeared to be stable at close to 0.5 SD (corresponding to approximately the 30th percentile) and 10 15% of patients were below the 5th percentile for height. In adults aged 40 y, height z scores approached zero (corresponding to the 50th percentile). However, mean weight z scores and percentages of ideal weight remained significantly lower in adults than in children with CF. There were no significant differences in mean z scores or the percentage of patients below the 5th percentile for height between the United States and Canada in the adult population. However, separation of weight and percentages of ideal weight curves between the United States and Canada was apparent in both sexes and Canada had significantly better weight z scores and percentages of ideal weights than the United States. Growth status in relation to diagnostic categories To identify clinical factors that may be associated with poor growth and a high prevalence of malnutrition, growth indexes were examined in subgroups of patients based on variables that reflect time of CF diagnosis, history of meconium ileus, pancreatic functional status, and CF genotype. The overall pattern of differences between the United States and Canada in each of the diagnostic categories is shown in Table 2. In general, patients with a new diagnosis of CF, those with meconium ileus, and those taking pancreatic enzyme supplements had poorer height and weight percentiles than those in the complementary categories. Lower values were also seen in those homozygous for F508 than in those with other genotypes. These patterns of differences were consistent within both countries. DISCUSSION The findings showed that children with CF in the United States (8) and Canada experience subnormal growth relative to NCHS and NHANES II reference standards, with mean height and weight at the 30 35th percentiles. Mean height and weight percentiles in children with CF in Canada were 4 5 percentiles higher than those in the United States, but percentages of ideal weight (104%) were similar in children from both countries. In adults with CF, mean heights were at the 37th percentile in both countries. However, the weight percentile and percentage of ideal weight were higher in Canada (weight: 26th percentile; percentage of ideal weight: 93%) than in the United States (weight: 21st percentile; percentage of ideal weight: 90%). In both countries, adults with CF had better height z scores but poorer weight z scores and percentages of ideal weight than did children with CF. These findings may reflect a survival advantage for those who achieved greater height. However, eventual weight loss in the later stages of the disease have been observed in this select group of survivors.

6 536 LAI ET AL FIGURE 3. Comparison of the prevalence of malnutrition by sex and age between patients with cystic fibrosis (CF) in the United States and Canada as a percentage of patients with height below the 5th percentile, with weight below the 5th percentile, and < 85% of ideal weight. For males, the number of observations are as follows: 3334 (0 2 y), 9278 (3 10 y), 6403 (11 17 y), and 8334 ( 18 y) for the United States and 431 (0 2 y), 1349 (3 10 y), 1041 (11 17 y), and 1494 ( 18 y) for Canada. For females, the number of observations are as follows: 2977 (0 2 y), 6313 (3 8 y), 7514 (9 17 y), and 6873 ( 18 y) for the United States and 401 (0 2 y), 957 (3 8 y), 1305 (9 17 y), and 1169 ( 18 y) for Canada. Significance was assessed by generalized estimating equations. Data are from the CF Patient Registries of the United States and Canada. Our analyses based on the CF Patient Registry data showed that the differences in growth of CF patients between the United States and Canada are substantially smaller than those reported by Corey et al (7), who compared CF patients from Boston and Toronto in the 1980s. In our study, height and weight were only 2 3 percentiles higher in male patients in Canada than in males in the United States, whereas Corey et al reported a 9-percentile difference for both height (Boston: 33rd percentile; Toronto 42nd percentile) and weight (Boston: 35th percentile; Toronto: 43rd percentile) between the Toronto and Boston groups. In female patients, height and weight were 4 percentiles higher in Canada than in the United States, whereas height was 11 percentiles higher (Boston: 33rd percentile; Toronto: 44th percentile) and weight was 2 percentiles lower (Boston: 31st percentile; Toronto: 29th percentile) in Toronto than in Boston. Two factors should be considered in the interpretation of these differences. First, the 2 studies used different standards of growth. The Toronto-Boston study used the British growth standards of Tanner et al (29), whereas the present study used NCHS (17) and NHANES II (20) standards. Second, growth of patients in Boston and Toronto may not represent that in the respective countries. NCHS and NHANES II standards for height in both sexes and for weight in males are generally higher than British standards; therefore, computed percentiles would be lower when NCHS and NHANES II standards are used. For example, height was 6 percentiles less and weight 12 percentiles less in Canada when NCHS and NHANES II standards were used than when British standards were used. In contrast, the opposite pattern was observed for weight standards for females. British standards are higher and thus the computed percentiles are lower when the British standards are used than when NCHS and NHANES II standards are used. These phenomena were described in a recent abstract (30). Therefore, lower growth percentiles in the present study than in Corey et al s (7) study in the respective countries largely reflect the use of different growth standards. In addition, because of the regular pattern of discrepancy between standards for height and weight in males, we concluded that the differences between the 2 countries in the present study were smaller than the regional differences in the earlier study (7). Note that weight percentiles were markedly higher in males than in females with CF when British standards were used (Boston: males, 35th percentile; females, 29th percentile), whereas only small sex differences were observed in children with CF when NCHS standards were used (United States: males, 30.4th percentile; females, 30.0th percentile). The opposite trend was observed for adult CF patients when NHANES II standards were used (United States: males, 19th percentile; females, 25th percentile). Such trends for sex differences were also observed in the Toronto and the Canadian populations. Several implications can be drawn from the aforementioned findings. First, the findings explain, at least in part, the inconsistent trend in sex differences as they relate to the growth of CF patients reported among studies (2, 5, 7, 8, 24, 31). For example, several studies (2, 5, 7, 31) have consistently shown better growth status, based on British percentiles, in males than in females with

7 GROWTH IN PATIENTS WITH CYSTIC FIBROSIS 537 TABLE 2 Growth status and prevalence of malnutrition by diagnostic characteristics in cystic fibrosis (CF) patients reported to the CF Patient Registries of the United States (US) and Canada Mean growth indexes Height Weight Percentage of ideal Height Weight <85% of Diagnostic category and country percentile percentile weight <5th percentile <5th percentile ideal weight % % % % Diagnosis New US (n = 2563) Canada (n = 323) Previous US (n = ) Canada (n = 7808) History of meconium ileus Yes US (n = 9572) , , , ,3 Canada (n = 1176) No US (n = ) Canada (n = 6955) Genotype F508/ F508 US (n = ) , , , , Canada (n = 2441) F508/other US (n = 8613) Canada (n = 1851) Other/other US (n = 562) Canada (n = 409) Pancreatic enzyme supplementation Yes US (n = ) , , , , Canada (n = 6739) No US (n = 2916) Canada (n = 676) Significantly different from the other US subgroup(s) within the diagnostic category, P < 0.05 (by generalized estimating equation methods that adjusted for repeated measures and stratified by sex). 2 Significantly different from the other Canadian subgroups within the diagnostic category, P < 0.05 (by generalized estimating equation methods that adjusted for repeated measures and stratified by sex). 3 Significantly different from Canada within the diagnostic category, P < 0.05 (by generalized estimating equation methods that adjusted for repeated measures and stratified by sex). CF. However, other studies (8, 24) reported opposite trends based on NCHS percentiles. Second, it indicates that the sex difference in weight (when expressed as percentiles or z scores) and in percentages of ideal weight observed previously in the CF population were at least partially dependent on the growth standards used for computation. In conjunction with the above argument, previous studies (2, 5, 7, 31) suggesting a possible link between higher mortality and poorer weight status for females than for males with CF warrants further examination. Use of other supplementary growth indexes that are independent of the influence of a reference population, such as gains and losses in height and weight as well as body mass index, may be essential for evaluating growth and its association with mortality in CF patients. Additionally, it is important to take into account differences in patients clinical characteristics, such as pancreatic status, pulmonary status, and genotype when examining the association between growth and survival in CF. Growth status of CF patients in the United States in the 1990s has improved considerably since the 1980s when the growth of patients from Toronto and Boston were examined. Median survival age increased from 26.9 to 31.3 y and the prevalence of malnutrition decreased by 20% in US patients from 1986 to 1996 (9). The median survival age of US CF patients is now virtually identical to that of Canadian patients, which was 31.0 y in 1995 (32). The significant improvement in nutritional status of CF patients in the United States over the past decade is due, at least in part, to the recognition of the importance of nutrition monitoring and to efforts made toward more aggressive nutrition interventions (12 14). Other factors may also play a role in the differences in nutritional status between CF patients in the United States and Canada. First, a lower prevalence of meconium ileus and apparent pancreatic insufficiency as well as a lower frequency of F508 mutations were noted in Canada than in the United States (Table 1). Adjustment for these factors resulted in a reduction in the significance

8 538 LAI ET AL (from P < to P < 0.01) of mean differences in growth indexes between the United States and Canada. Second, other clinical profiles such as pulmonary function and rate of respiratory infection were not determined; this may have contributed to the differences between the 2 countries observed in our study. Third, differences in health care delivery between the United States and Canada likely complicate growth comparisons as well. Whereas CF patients in Canada have enjoyed universal health insurance for many decades, 98.1% of CF patients in the United States were insured in 1994 (33). However, mildly affected patients may be underrepresented or use of supplemental nutrition may be lower in the US CF population because of health care expenses. In conclusion, the most striking findings of this study are the substantially smaller differences in growth indexes between the 2 countries compared with results documented in the 1980s from 2 regional CF centers (Boston and Toronto). However, direct comparison of mean growth percentiles from studies using different growth standards is inappropriate because systematic differences exist in growth standards, which affect, in particular, the comparison of growth between males and females. Although we observed diminishing differences in growth between the United States and Canada during the current decade, it is clear that Canadian children with CF have remained slightly taller and heavier for their age than US patients with CF. In addition, US and Canadian adults with CF have similar heights but Canadians are significantly heavier. Further studies examining the contributions of various clinical, genetic, and environmental factors associated with CF are needed to explain the observed differences in nutritional status between CF patients in the United States and Canada. This study was made possible by the efforts of all the clinic directors and coordinators in Canada and the United States, who make the national registries such a valuable resource for CF care and research. REFERENCES 1. Kosorok MR, Wei WH, Farrell PM. The incidence of cystic fibrosis. Stat Med 1996;15: Corey M, Farewell V. Determinants of mortality from cystic fibrosis in Canada, Am J Epidemiol 1996;143: Kraemer R, Rudeberg A, Hadorn B, et al. Relative underweight in cystic fibrosis and its prognostic value. Acta Paediatr Scand 1978;67: Soutter VL, Kristidis P, Gruca MA, et al. Chronic undernutrition/growth retardation in cystic fibrosis. Clin Gastroenterol 1986;15: Kerem E, Reisman J, Corey M, et al. Prediction of mortality in patients with cystic fibrosis. N Engl J Med 1992;326: Zemel BS, Riley EM, Stallings VA. Evaluation of methodology for nutrition assessment in children: anthropometry, body composition, and energy expenditure. Annu Rev Nutr 1997;17: Corey M, McLaughlin FJ, Williams M, et al. A comparison of survival, growth, and pulmonary function in patients with cystic fibrosis in Boston and Toronto. J Clin Epidemiol 1988;41: Lai HC, Kosorok MR, Sondel SA, et al. Growth status in children with cystic fibrosis based on the national Cystic Fibrosis Patient Registry data: evaluation of various criteria used to identify malnutrition. J Pediatr 1998;132: Cystic Fibrosis Foundation. Cystic Fibrosis Foundation Patient Registry Annual Data Report. Bethesda, MD: Cystic Fibrosis Foundation, Crozier DN. Cystic fibrosis a not so fatal disease. Pediatr Clin North Am 1974;21: Shwachman H. Cystic fibrosis. In: Kendig EL, Chernick V, eds. Disorders of the respiratory tract in children. Philadelphia: WB Saunders, 1977: Ramsey BW, Farrell PM, Pencharz P, et al. Nutritional assessment and management in cystic fibrosis: a consensus report. Am J Clin Nutr 1992;55: Cystic Fibrosis Foundation. Clinical practice guidelines for cystic fibrosis. Bethesda, MD: Cystic Fibrosis Foundation, Cannella PC, Bowser EK, Guyer LK, Borum PR. Feeding practices and nutrition recommendations for infants with cystic fibrosis. J Am Diet Assoc 1993;93: FitzSimmons SC. The changing epidemiology of cystic fibrosis. J Pediatr 1993;122: Dean AG, Dean JA, Coulombier D, et al. Epi Info, version 6: a word processing, database and statistical program for epidemiology on microcomputers. Atlanta: Centers for Disease Control and Prevention, Hamill PVV, Drizd TA, Johnson CL, Reed RB, Roche AF, Moore WM. Physical growth: National Center for Health Statistics percentiles. Am J Clin Nutr 1979;32: Dibley MJ, Goldsby JB, Staehling NW, Trowbridge FL. Development of normalized curves for the international growth reference: historical and technical considerations. Am J Clin Nutr 1987;46: Dibley MJ, Staehling NW, Nieburg P, Trowbridge FL. Interpretation of Z-score anthropometric indicators derived from the international growth reference. Am J Clin Nutr 1987;46: US Department of Health and Human Services. Anthropometric reference data and prevalence of overweight. United States, Data from the National Health Survey. Hyattsville, MD: US DHHS. [Series 11, no. 238, DHHS publication no. (PHS) ] 21. National Research Council. Recommended dietary allowances. 10th ed. Washington, DC: National Academy Press, Demko CA, Byard PJ, Davis PB. Gender differences in cystic fibrosis: Pseudomonas aeruginosa infection. J Clin Epidemiol 1995;48: Britton J. Effects of social class, sex and region of residence on age at death from cystic fibrosis. BMJ 1989;298: Rosenfeld M, Davis R, FitzSimmons S, Pepe M, Ramsey B. Gender gap in cystic fibrosis mortality. Am J Epidemiol 1997;145: Fleiss JL. Statistical methods for rates and proportions. New York: John Wiley and Sons, Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73: Karim MR, Zeger SL. GEE: a SAS macro for longitudinal data analysis (version 1). Baltimore: Department of Biostatistics, The Johns Hopkins University, (Technical report no. 674.) 28. Su JQ, Wei LJ. Nonparametric estimation for the difference or ratio of median failure time. Biometrics 1993;49: Tanner JM, Whitehouse RH, Takaishi M. Standards from birth to maturity for height, weight, height velocity and weight velocity. British children, Arch Dis Child 1966;41: Shin J, Corey M, Kalnins D, Pencharz P. A comparison of anthropometric standards for evaluation of CF patients. Pediatr Pulmonol Suppl 1997;14:A390 (abstr). 31. Corey ML. Longitudinal studies in cystic fibrosis. In: Sturgess JM, ed. Perspectives in cystic fibrosis: Proceeding of the Eighth International Congress on Cystic Fibrosis. Mississauga, Canada: Imperial Press, 1980: Canadian Cystic Fibrosis Foundation. Canadian Patient Data Registry National Report. Toronto, Canada: Canadian Cystic Fibrosis Foundation, Cystic Fibrosis Foundation. Cystic Fibrosis Foundation Patient Registry Annual Data Report. Bethesda, MD: Cystic Fibrosis Foundation, 1994.