 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* Department of Biostatistics/Medical Informatics, University of Wisconsin; Madison, Wisconsin
Department of Pediatrics, University of Wisconsin; Madison, Wisconsin
Department of Nutritional Sciences, University of Wisconsin; Madison, Wisconsin
|| Department of Statistics, University of Illinois at Urbana-Champaign, Urbana-Champaign, Illinois
¶ Department of Pediatrics, Medical College of Wisconsin, Milwaukee, Wisconsin
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods. Cognitive assessment data were obtained for 89 CF patients (aged 7.3-17 years) during routine clinic visits. Patients had been enrolled in either the screened (N = 42) or traditional diagnosis (control) group (N = 47) of the Wisconsin CF Neonatal Screening Project. The Test of Cognitive Skills, Second Edition was administered to generate the Cognitive Skills Index (CSI) and cognitive factor scores (Verbal, Nonverbal, and Memory).
Results. Cognitive scores in the overall study population were similar to normative data (CSI mean [standard deviation]: 102.5 [16.6]; 95% confidence interval: 99.1-105.9). The mean (standard deviation) CSI scores for the screened and control groups were 104.4 (14.4) and 99.8 (18.5), respectively. Significantly lower cognitive scores correlated with indicators of malnutrition and unfavorable family factors such as single parents, lower socioeconomic status, and less parental education. Our analyses revealed lower cognitive scores in patients with low plasma 
-tocopherol (-T) levels at diagnosis. In addition, patients in the control group who also had vitamin E deficiency at diagnosis (-T < 300 µg/dl) showed significantly lower CSI scores in comparison with -T–sufficient control subjects and both deficient and sufficient -T subsets of screened patients.
Conclusion. Results suggest that prevention of prolonged malnutrition by early diagnosis and nutritional therapy, particularly minimizing the duration of vitamin E deficiency, is associated with better cognitive functioning in children with CF.
Key Words: cognitive function • cystic fibrosis • malnutrition • neonatal screening • vitamin E deficiency
Abbreviations: CF, cystic fibrosis • SES, socioeconomic status • -T, -tocopherol • TCS/2, Test of Cognitive Skills, Second Edition • CSI, Cognitive Skills Index • MI, meconium ileus • PS, pancreatic sufficient • PI, pancreatic insufficient
It is commonly assumed that cognitive development is not altered in children with cystic fibrosis (CF).1,2 Most patients with CF, however, are at high risk for malnutrition, particularly during the prediagnosis, symptomatic phase of the disease in which pancreatic insufficiency leads to inadequate absorption of fat, protein, and fat-soluble vitamins.3–5 Previous studies of cognitive development in non-CF children have not always taken into account the nutritional status of the participants,5–7 although early nutritional insult is one of many factors associated with stunted cognitive development.8,9 Other variables associated with cognitive development include socioeconomic status (SES), mothers’ and fathers’ education, and parents’ marital status.10 The interplay of these variables on cognitive performance is not well understood,8 particularly in populations that may be at high risk for compromised nutritional development during the first few years of life when brain development is critical.8,9
As part of the Wisconsin Cystic Fibrosis Neonatal Screening Project,5,11 we were presented with a unique opportunity to evaluate cognitive function as related to age of diagnosis and nutritional status. Although this was not one of the objectives when the investigation was planned in 1984, we recognized the importance of cognitive assessment when we found to our surprise that protein-energy malnutrition was so severe with delayed diagnosis that head circumference percentiles in the control group were significantly lower than in the screened group.12 In addition, the occurrence of deficiencies in micronutrients such as vitamin E13,14 was common and severe enough to raise concern about central nervous system effects.8,15,16 Therefore, the present study aims included the following: 1) to confirm that cognitive functioning of children with CF overall is not different from normative samples; 2) to investigate the relationship between early indicators of malnutrition and long-term cognitive development of children with CF; and 3) to investigate whether early diagnosis of CF via neonatal screening results in better cognitive development than delayed diagnosis via symptoms, controlling for socioeconomic factors and evidence of early malnutrition.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
F508 mutation was used as an improved screening procedure thereafter until randomization was concluded on June 30, 1994. Children in the screened group who had positive screening tests were contacted immediately for follow-up sweat-test evaluation at 1 of Wisconsin’s 2 CF centers. For avoiding selection bias that could result from missed cases in the traditional diagnosis group, trypsinogen results in the control group were unblinded as children reached 4 years of age.16 Any children who had positive screens and had not already received a diagnosis of CF were then contacted for follow-up evaluation. Once the diagnosis was made by a sweat chloride level of 
60 mEq/L and patients were enrolled by parental consent, all patients were treated longitudinally with the same evaluation and treatment protocol17 designed to ensure optimal care while providing quantitative outcome assessments. The protocol for nutritional therapy included pancreatic enzyme supplements, increased caloric intake at 120% to 150% of the Recommended Dietary Allowance,12 fatty acid supplements in the form of Lipomul or corn oil margarine when linoleate levels were <26%, and water miscible multivitamin supplements. The doses for vitamin E supplements were 50 IU/day for infants, 100 IU/day for children 1 to 10 years of age, and 200 IU/day for those older than 10 years. Plasma -tocopherol (-T) levels were closely monitored until they rose to normal13,19; thereafter, levels were measured annually. Table 1 describes the patients involved in this study, who ranged in age from 7.3 to 17 years. Children for cognitive assessment were recruited from participants in the overall Wisconsin Cystic Fibrosis Neonatal Screening Project11,12,17 on the basis of the following eligibility criteria: 1) enrolled before March 1999 (ie, positive sweat test before March 1999), 2) attained 7 years of age before August 2002, 3) currently receiving care at 1 of the 2 CF centers collaborating in the Screening Project, and 4) parents provided consent for child’s participation. Randomization of 650 341 newborns led to identification of 160 children with CF and 138 enrollees, 73.2% of whom met the eligibility criteria for the cognitive study. Of the 101 eligible patients, consent to participate was obtained for 93 (92.1%), and cognitive assessment data were obtained for 89 (88.1% of those eligible). The Human Subjects Committee at the University of Wisconsin and the Research and Publications Committee and Human Rights Board at Children’s’ Hospital of Wisconsin approved the randomization, evaluation, and treatment methods, as well as cognitive studies.
|
Normative tables for the TCS/2 provided percentiles for all cognitive outcomes on the basis of student age, to the nearest month. In addition, normative tables provided age-based standard scores for CSI (mean: 100; standard deviation [SD]: 16; n = 70 833 students). Standard scores with the same mean and SD as CSI were obtained for the cognitive abilities factors by converting the percentiles from the normative tables into standard scores using Splus’ qnorm function.21
Independent Variables
In the general pediatric population, cognitive development may be influenced by multiple factors, including but not limited to nutrition, SES, and family factors.10 In addition, cognitive development in children with CF could potentially be associated with disease- and screening study–specific variables (eg, study group, genotype, treatment center, history of meconium ileus [MI]). The variables used in this study are summarized below.
F508 and b) heterozygous F508 or 2 non-F508 CF mutations. Pancreatic status was designated as pancreatic sufficient/probably pancreatic sufficient (PS) or pancreatic insufficient/probably pancreatic insufficient (PI), as determined and described previously.4 Lung disease was assessed by pulmonary function tests and chest radiographs obtained at or near the time of cognitive assessment.22 -T), measures of essential fatty acid status25 (linoleic acid as a percentage of total fatty acids [%18:2] and the triene/tetraene ratio), and serum albumin14 levels. The anthropometric variables were obtained from the time of diagnosis, 3 months after diagnosis, and at the time of cognitive assessment. Biochemical measures were obtained at the time of diagnosis, and -T levels were also examined at or near the time of cognitive assessment. Selected anthropometric and biochemical measures were also summarized as dichotomous variables using the following groupings: <10th percentile for height, weight, and head circumference; -T of <300 µg/dL for vitamin E status26,27; retinol of <20 µg/dL for vitamin A status14,28; and <26% linoleic acid for essential fatty acid deficiency.25 Height and weight gains in percentiles from diagnosis to 3 and 6 months after diagnosis were also calculated.
A final study-specific variable, referred to as group/-T, was created by categorizing the screened and control groups into the following 4 subgroups on the basis of plasma -T levels at diagnosis: 1) control patients, -T <300 µg/dL (C < 300E); 2) control, -T 300 µg/dL (C 300E); 3) screened, -T <300 µg/dL (S < 300E); and 4) screened, -T 300 µg/dL (S 300E). This grouping was based on 1) observations on the relationship of plasma -T levels in CF and demonstrated biological antioxidant deficiency,29 2) literature that links early vitamin E deficiency to compromised physiologic outcomes,8,13,29,30 and 3) univariate analyses herein that identified -T <300 µg/dL at diagnosis as a strong predictor of subsequent cognitive functioning in this CF sample.
Statistical Analyses
T tests or 2-sample Wilcoxon tests were used to compare the equality of continuous variables between 2-level independent variables; test selection depended on the distribution of the data. The 1-sample t test was used to compare observed values to an expected mean. 
2 tests were used to test the relationship between dichotomous variables. Correlations were calculated to examine the relationship between pairs of variables (Pearson for normally distributed data; Spearman otherwise). Children with MI were omitted from analyses that evaluated the effect of screening because of the possibility that including those with MI in the control group would bias the results.
Stepwise regression was used to examine the group/-T effects on cognitive development, controlling for other known or potential correlates of cognitive function. Variables considered for inclusion, using entry/exit P = 0.15, were group/-T status, family/SES characteristics, anthropometric and biochemical indices at diagnosis, pancreatic status, genotype, CF center, gender, and age at diagnosis. Using Scheffe procedure for multiple comparisons,31 contrasts between group/-T levels were tested at the same significance level as the overall test of group/-T only when the overall test was rejected. Model validity was determined by examining residuals. All statistical analyses were completed using SAS32 and Splus23 software and 2-tailed tests of significance.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Demographic and clinical comparisons of the 71 screened and control patients revealed group differences similar to those reported previously for other evaluations with larger samples.4,12 The age of diagnosis differed (P < .001) with the screened group at mean (SD) of 16.3 (46.0) weeks compared with 122 (119) weeks, which is similar to national data for traditional diagnoses during 1985-1994.33 In addition, there were more control subjects who were PS compared with the screened group (32% vs 11%; P = .03). Despite the nearly 3-fold higher prevalence of intact pancreatic function, the control group showed significantly worse nutritional status based on anthropometric indexes with lower height-for-age z scores at diagnosis (P = .019) and at both 3 months (P = .016) and 6 months (P = .001) after diagnosis. In this sample subset, however, there were no significant differences in weight or head circumference, which contrasts with findings in the larger screened and control groups published previously.4,12
Biochemical measures of nutritional status at diagnosis did not differ significantly between groups as reported before,4 but the duration of nutrient deficiencies was obviously longer in the control patients with PI. At the time of diagnosis, screened non-MI patients with pancreatic insufficiency at 16.9 weeks’ average age had mean (SD) plasma -T levels of 424 (336) µg/dL, and the traditional diagnosis counterparts at 77.4 weeks were 280 (268) µg/dL. In contrast, the corresponding groups with pancreatic sufficiency were 885 (443) and 832 466) µg/dL, respectively, when diagnosed at 10.8 and 214 weeks of age. Overall, 49% of these patients had -T levels <300 µg/dL at diagnosis, but only 8% and 3% were low 3 and 6 months after treatment, respectively, and no patient with vitamin E deficiency at diagnosis was low after 6 months on the evaluation and treatment protocol. Observations on vitamin A status were somewhat similar in that 34% of the patients were low at diagnosis, and none had plasma retinol levels <20 µg/dL when assessed 6 months later. In addition, there were no group differences in the variables representing SES, parents’ education, or marital status.
A 1-sample t test of the CSI scores indicated that the 89 CF patients did not differ significantly from the general population mean of 100 (CSI mean [SD]: 102.5 (16.6); P = .15; 95% confidence interval for CSI 99.1-105.9). Similarly, the signed-rank tests of the cognitive factor percentiles showed no significant differences between the CF and norming samples. In addition, analysis of variance testing the equality of scores across TCS/2 levels revealed no significant differences. This finding is consistent with the TCS/2 reports that scores across levels (and student ages) are comparable.20
Factors Associated With Cognitive Development of Children With CF
Comparison of the non-MI screened group and the traditional diagnosis control group did not reveal any significant differences. The mean (SD) CSI scores were 104.4 (14.4) and 99.8 (18.5), respectively (P = .24). Results were similar for the 3 cognitive factors. Specifically, means (SD) for screened versus controls were 104.6 (12.6) versus 101.8 (17.9) for verbal (P = .45), 106.4 (16.5) versus 100.7 (18.3) for nonverbal (P = .17), and 100.3 (15.1) versus 95.4 (19.5) for memory (P = .23).
Our results indicated that children with poorer nutritional status at diagnosis and both 3 and 6 months after diagnosis scored lower on cognitive assessments than children with better nutritional status at these early time points. Figure 1 depicts cognitive scores in relationship to anthropometric measures of nutritional status. Significant differences were found between groups with evidence of malnutrition, ie, patients below compared with those above the 10th percentile for the following variables (and outcomes at specified time points): weight (at diagnosis: CSI, P = .04; and Memory, P = .01; at 3 months after diagnosis: CSI, P = .05, Nonverbal, P = .02, and Memory, P = .003) and height (at diagnosis: Nonverbal P = .05; at 3 months after diagnosis: Nonverbal P = .04). After 6 months of therapy, only 1 significant difference in cognitive scores was observed between those below and above the 10th percentile for the anthropometric measures of nutritional status. Specifically, children with height lower than the 10th percentile 6 months after diagnosis (n = 10) had mean (SD) Memory scores of 86.1 (18.5) compared with mean (SD) of 99.9 (16.9) in the 55 children with height at or above the 10th percentile (P = .02). Significant differences in cognitive outcomes for low versus high percentile groupings were not found for head circumference at any of the 3 time points.
|
-T were found for 3 of the 4 cognitive outcomes (CSI, P = .006; Verbal, P = .03; Nonverbal, P = .01). Differences in cognitive scores between low and high vitamin A groupings were observed for 1 cognitive outcome (Verbal, P = .04), but there were no significant differences related to low versus adequate linoleic acid levels. Results (not depicted) also indicated that there were no significant differences in cognitive performance for the following comparisons: MI versus non-MI, PS versus PI, F508/ F508 versus F508/CF or CF/CF, CF centers (Madison vs Milwaukee), and male versus female.
|
Correlations between cognitive outcomes and continuous or multilevel ordinal categorical variables are summarized in Table 2. Nutritional measures that were obtained at diagnosis and correlated significantly with 1 or more cognitive outcomes included weight-for-age z score (CSI, r = .32; Nonverbal, r = .29; Memory, r = .30), height-for-age z score (CSI, r = .26; Nonverbal, r = .27; Memory, r = .24), plasma -T (CSI, r = .28; Nonverbal, r = .27), and serum albumin (Verbal, r = .30). Other variables that were obtained at the time of diagnosis and correlated significantly with 1 or more cognitive outcomes included family income (Memory, r = .25), mother’s education (CSI, r = .32; Nonverbal, r = .33), and father’s education (CSI, r = .24; Verbal, r = .25; Nonverbal, r = .24).
|
Results for the 4 cognitive outcomes are shown in Fig. 3 by study group and vitamin E status at the time of diagnosis (group/-T categories). Adjusted means on the basis of the corresponding analysis of covariance analyses are summarized in Table 3. Before adjusting for covariates, CSI differed significantly for group/-T (P = .016) as did Memory (P = .032). The comparison of screened versus control groups, ie, combining the -T subgroups, was not significant for either of these cognitive factors. However, the significant pairwise contrasts for CSI included C < 300E versus C 300E and C < 300E versus S 300E; for Memory, these and C < 300E versus S < 300E were significant. After adjusting for the outcome-specific covariates (income, mother’s and father’s education, marital status, and any variables identified by the stepwise regression), all but Memory differed significantly across group/-T levels. Pairwise contrasts indicated that the C < 300E group (vitamin E–deficient control group) consistently had lower average scores on all 4 cognitive outcomes than each of the other group/-T levels. For example, after adjusting for the previously mentioned variables, the C < 300E group was significantly lower in terms of global cognitive abilities (CSI) than each of the other 3 subgroups (pairwise differences ranged from 12.5 to 16.2 points).
|
|
-T levels. Both control subgroups differed significantly from the screened subgroups for age of diagnosis (P < .05). The 2 control groups also differed significantly in age of diagnosis, ie, those with -tocopherol 300 µg/dl had a greater delay in diagnosis; this delay was associated with less pancreatic insufficiency. |
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The Wisconsin trial4,12 and many other studies3,4,37 have shown that fat-soluble vitamin deficiencies are very common in children with CF and may be difficult to correct.37 For instance, Farrell et al27 showed that vitamin E deficiency was universally present in CF patients who have pancreatic insufficiency and do not receive tocopherol supplements. These abnormalities develop very early in the life of a CF patient with pancreatic insufficiency according to data published from the Colorado3,4,37 and Wisconsin5,12 newborn screening studies. Thus, we assume that in our patients, vitamin E deficiency began in those with pancreatic insufficiency shortly after birth and that the age of diagnosis provides an estimate of the duration of low -T levels. Ferarchak et al37 also found multiple vitamin deficiencies, as did the Wisconsin investigators,12 but encountered difficulty correcting vitamin E deficiency with tocopherol supplements (possibly because of low doses and/or compliance problems). With doses described herein and elsewhere13,19 and regular counseling, we rapidly corrected vitamin E deficiency in almost all of the patients.
When the level of plasma -T is <300 µg/dL, biological antioxidant insufficiency may be assumed on the basis of data regarding development of membrane instability with accelerated cell (eg, erythrocyte) destruction both in vitro38,39 and in vivo.28,30 As vitamin E deficiency continues, severe hemolytic anemia13 and nervous system abnormalities16,40 may ensue. This is consistent with findings in laboratory animals, such as chicks, in which oxidative destruction of neurons occurs in the cerebellum and to a lesser extent in the cerebrum.16 Recognizing the risks of fat-soluble vitamin deficiencies, we organized the Wisconsin trial with an evaluation and treatment protocol that would ensure rapid correction of and sustained prevention of vitamin E deficiency. However, this rapid correction could be applied only to those whose CF is diagnosed early through screening, and the control subjects with pancreatic insufficiency had an average of 107 weeks of vitamin E deficiency. Consequently, antioxidant deficiency in these patients occurred during a critical phase of brain development.
A variety of studies involving non-CF children have shown that cognitive development may be influenced by many factors, including family conditions8,10 and early nutritional status.8,9 For example, in a study of extremely low birth weight infants, those with early malnutrition (head circumference percentile <3% at 4, 8, or 12 months of age) scored significantly worse on a measure of cognitive ability than their better nourished extremely low birth weight peers.41 Similarly, a meta-analysis of the effects of breastfeeding on cognitive development42 showed that breastfed infants consistently scored higher than formula-fed peers after adjusting for covariates such as parental education and family income. This benefit has been linked to the superior fatty acid composition of human milk compared with infant formulas.43 Given the potential association between early malnutrition and subsequent cognitive development in the general pediatric population, it is reasonable to hypothesize that children with CF, particularly those who are at risk for fat-soluble vitamin malabsorption, may experience cognitive deficits from prolonged severe malnutrition and cognitive benefits from early diagnosis. The results of this study support that hypothesis. Specifically, the children with traditional delayed diagnosis (ie, control subjects) and vitamin E deficiency (as indicated by -T <300 µg/dL) performed worse (P < .05) on cognitive outcomes than both control subjects with adequate vitamin E levels and patients whose CF was diagnosed early through screening (even if the screened patients had short-term vitamin E deficiency). The unadjusted discrepancies between the low -T/control group and the other groups ranged from 9.4 to 15.8 points on the CSI; these differences are clinically as well as statistically significant; the unadjusted differences for each cognitive factor ranged from 8.8 to 13.1, 6.3 to 14.6, and 12.5 to 16.5 points for the Verbal, Nonverbal, and Memory factors, respectively.
There are some limitations in this study. First, comparison of the participants with the nonparticipants showed that those who were assessed with cognitive testing had a lower percentage of nonwhite children (2.3% vs 12.8%). Thus, the differences that we report here are most pertinent to white children with CF (who represent 95.3% of CF patients in the United States40). Additional study of minority patients would be needed to determine whether these results apply across a wider range of ethnicities. Second, more precise and/or additional covariates (income estimates and parent’s mental health status, respectively) might improve the modeling of cognitive development in this population. Third, the failure to find significant differences between the screened and control groups may be attributable to the existence of a higher percentage of PS patients in the control group (and subsequently milder disease and less risk of malnutrition). Fourth, it is possible that there are other explanations for the decreased cognitive performance among patients with delayed diagnosis and prolonged vitamin E deficiency. For example, one might argue that the children whose CF was diagnosed early had increased parental stimulation as a result of starting CF treatment earlier than their control counterparts. If this argument explained a significant amount of the cognitive score variations, then we would not expect to see a difference between the low and adequate -T control subgroups, whereas those differences are consistently the largest or second largest found across 4 cognitive outcomes. Finally, other factors that were not studied may influence cognitive function. However, we have examined potential confounders and nothing alters our conclusions. For example, when the analyses were rerun after removing children with attention-deficit/hyperactivity disorder, the differences between C < 300E and the other groups became larger. The same was true when the subset removed consisted of patients whose CF was diagnosed because they had an older sibling with CF.
In conclusion, we recognize that this study is the first to examine and identify a potential association among time of diagnosis, vitamin E status, and subsequent cognitive development in children with CF, as well as the first systematic assessment of cognitive ability in CF patients. The observation that we report of a screening/-T interaction linking vitamin E deficiency with cognitive outcomes has profound implications for children with CF. Our findings, however, are consistent with other observations revealing a link between vitamin E deficiency and physiologic functions that are sensitive to antioxidant deficiency.30 Low vitamin E levels may cause hemolytic anemia, which during early development has been implicated as a potential factor in stunted cognitive development.8 Vitamin E deficiency also causes neuronal degeneration in laboratory animals16 and humans44–47 with prolonged periods of low tocopherol levels. In addition, a growing body of evidence from Alzheimer research has revealed an association between vitamin E and cognitive performance decline.48–51
Our results, therefore, add to the evidence linking vitamin E with neurodevelopment and cognitive function. The screening/vitamin E interaction also adds important information to the assessment of benefits and risks of CF neonatal screening and to the debate over whether to recommend this routine diagnostic method. It should be clear that nonscreened patients who have CF that is diagnosed by traditional methods do not have access to the potential cognitive benefit associated with preventing prolonged malnutrition. This opportunity difference applies to at least half of the CF patients: those at risk for fat-soluble vitamin deficiencies. To summarize, in addition to better nutritional outcomes,4,12,52 we have provided evidence of potential cognitive benefits associated with better nutritional status, which can be achieved through early diagnosis and preventive therapy with diet, pancreatic enzymes, and fat-soluble vitamin supplements.5,12–14,17 This strengthens the argument for universal CF neonatal screening.53
|
ACKNOWLEDGMENTS |
|---|
We thank M. Durkin for advice on this study and helpful suggestions on the manuscript. We remain deeply grateful to the following investigators, who have participated in the Wisconsin Cystic Fibrosis Neonatal Screening Project: University of Wisconsin, Madison, Medical School, Madison: N. Fost, E. Mischler, M. Palta, A. Tluczek, M. Block, L.A. Davis, L. Feenan, D. Pfeil, K. Moucha, L.J. Wei, BS Wilfond, A. von Egmond, J. Sharp, L. Loveland, R. Brown; Medical College of Wisconsin, Milwaukee: W.T. Bruns, H. Colby, W. Gershan, C. McCarthy, L. Rusakow, M.E. Freeman, K. Riley; and Wisconsin State Laboratory of Hygiene, Madison: G. Hoffman, D.J. Hassemer, and R.H. Laessig.
|
FOOTNOTES |
|---|
Reprint requests to (P.M.F.) Department of Pediatrics, University of Wisconsin Medical School, Rm 1217 MSC, 1300 University Ave, Madison, WI 53706-1532. E-mail: pmfarrel{at}facstaff.wisc.edu
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Steinraths, H. D. Vallance, and A. G. F. Davidson Delays in diagnosing cystic fibrosis: Can we find ways to diagnose it earlier? Can Fam Physician, June 1, 2008; 54(6): 877 - 883. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. M Balfour-Lynn Newborn screening for cystic fibrosis: evidence for benefit Arch. Dis. Child., January 1, 2008; 93(1): 7 - 10. [Full Text] [PDF]
|
|
|
|
 |
|
 R. Padman, S. A. McColley, D. P. Miller, M. W. Konstan, W. J. Morgan, M. S. Schechter, C. L. Ren, J. S. Wagener, and for the Investigators and Coordinators of the Epid Infant Care Patterns at Epidemiologic Study of Cystic Fibrosis Sites That Achieve Superior Childhood Lung Function Pediatrics, March 1, 2007; 119(3): e531 - e537. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Comeau, F. J. Accurso, T. B. White, P. W. Campbell III, G. Hoffman, R. B. Parad, B. S. Wilfond, M. Rosenfeld, M. K. Sontag, J. Massie, et al. Guidelines for Implementation of Cystic Fibrosis Newborn Screening Programs: Cystic Fibrosis Foundation Workshop Report Pediatrics, February 1, 2007; 119(2): e495 - e518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. D. Grosse, C. A. Boyle, A. Kenneson, M. J. Khoury, and B. S. Wilfond From Public Health Emergency to Public Health Service: The Implications of Evolving Criteria for Newborn Screening Panels Pediatrics, March 1, 2006; 117(3): 923 - 929. [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Wagener States urged to consider newborn screening for cystic fibrosis AAP News, September 1, 2004; 25(3): 124 - 125. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
