What Is Known
- Failure to thrive is a common diagnosis with the majority of children having nonorganic etiologies of their poor growth.
- Laboratory evaluation is rarely helpful in determining the etiology of a patient's failure to thrive.
What Is New
- Fewer tests are being performed in children with FTT than previously reported but the diagnostic yield of testing remains low.
- In children with FTT, standard behavioral and nutritional interventions result in weight gain in the majority of children.
- In children with nonorganic FTT, adherence to behavioral interventions is associated with improved weight.
Failure to thrive (FTT) occurs when a child's rate of weight gain is below expectations based on age- and sex-specific growth charts (1). Although data regarding the prevalence of FTT in the United States are lacking, a 2006 Danish cohort study demonstrated a prevalence of 1.2% to 22.2% depending on the diagnostic criteria used (2). Large studies of FTT published in the 1970s and 1980s, mostly of hospitalized patients, have shown that the majority of children with FTT do not have organic disease and that without suggestive findings on history or physical examination, laboratory evaluation of these children is not useful (3–9). In the intervening decades, endoscopy has become increasingly used as a diagnostic modality, celiac disease has become more prevalent, and eosinophilic esophagitis has emerged as a distinct pathological entity. Although many studies have examined FTT in the inpatient setting, fewer have described FTT in the outpatient setting (3,8,9). In addition, none of these studies has examined FTT referrals solely in a pediatric gastroenterology (GI) clinic.
When a child presents with FTT, the physician must determine if the FTT is due to a symptom of underlying organic pathology, is nonorganic, or a combination of both. It has been suggested that the terms organic and nonorganic FTT be abandoned (1). Although many children may have both organic and nonorganic factors that contribute to their FTT, determining the primary etiology guides treatment for FTT. When organic disease is present, treatment focuses on the underlying pathology; when organic disease is absent, treatment focuses on behavioral and environmental modifications to optimize nutrition. Although children with organic disease often benefit from behavioral interventions in addition to management of the underlying organic disease, behavioral interventions alone are unlikely to lead to weight gain. Hence, distinguishing between organic FTT and nonorganic FTT was important to our study.
In the present study, we aimed to prospectively describe the clinical characteristics, diagnostic workup, interventions, and outcomes of children referred to a tertiary care pediatric GI clinic with the diagnosis of FTT.
We prospectively enrolled all English-speaking children seen for the first time between July 2013 and September 2014 in the outpatient pediatric GI clinic for evaluation of FTT via continuous enrollment; 1 family declined enrollment. Children were diagnosed with FTT if they met 1 of the 7 standard diagnostic criteria: weight <75% of median weight for age, weight <80% of median weight for length, body mass index (BMI) for age <5th percentile, weight for age <5th percentile, length for age <5th percentile, weight deceleration defined by crossing >2 major percentile lines for weight from birth, or conditional weight gain in the lowest 5%, adjusted for regression toward the mean from birth until weight within the given age group using the WHO growth velocity standards (2,10). We also categorized children as being malnourished based on z score, weight gain, and intake criteria from the American Society for Parenteral and Enteral Nutrition (11). We did not calculate change in weight-for-length and BMI z scores because of uncertain accuracy of heights/lengths before referral to our clinic. Children were plotted on standard WHO growth charts for children younger than 2 years and CDC growth charts for children of 2 years or older. We obtained BMI for all children and weight-for-length/height z scores for children younger than 5 years so that we were able to compare children younger than 2 years and those older than 2 years and so children who crossed 2 years of age during the study period could be analyzed. Clinical judgment was also used to ensure that the individual child was truly failing to thrive (ie, children who were felt to have constitutional delay of growth or were small for gestational age and following their own curve were not considered to be FTT) (12). Informed consent was obtained from parents and assent was obtained from patients when appropriate.
Historical information collected included demographic data and medical and social history. Standard anthropometric data including weight, length (<2 years of age) or height (≥2 years of age), BMI, and occipital frontal circumference (OFC) for children younger than 2 years were also obtained. Values were converted to z scores using the WHO Anthro program (13). Weights (using the Scale-Tronix 4802 Pediatric/Infant scale for children <2 years and Scale-Tronix 5002 for those ≥2 years (Wheaton, IL)), lengths (using the Harpenden Infant Measuring Table), heights (using the Harpenden Stadiometer (Crosswell, UK)), and OFCs (using standard paper tape measures) were obtained during routine clinical care in the pediatric gastroenterology clinic by personal trained to perform anthropometric measurements. For premature infants, corrected age was used until 2 years of age.
The primary pediatric gastroenterologist determined the workup and recommendations for treatment. Information and results of the workup of FTT including laboratory, imaging, and endoscopic evaluations were obtained. We also obtained results of laboratory tests done by the primary care provider. In addition, behavioral and nutritional recommendations for treating the child's FTT were obtained. Interventions were determined by the treating gastroenterologist but standard interventions included increasing caloric density of breast milk, infant formula, cow milk, or food; limiting mealtime or duration of feeds; structuring meals and snacks with techniques such as family meals and no distractions during meals; providing 3 meals and 1 to 2 snacks a day or feeding at appropriate intervals for infants; no grazing between meals and snacks (ie, having the child be feed at specified intervals with no access to food or calorie containing beverages in between); no juice or other sugar-sweetened beverages; and multivitamin supplementation.
Laboratory and examination charges were calculated based on standard charges at the Children's Hospital of Wisconsin irrespective of where the child had testing performed and excluded any genetic testing because of the variability in costs for these tests. Physician fees were also excluded. When counting individual tests, a complete blood count, basic metabolic panel, albumin, thyroid screen (TSH and Free T4), celiac screen (serum IgA level and anti-tissue transglutaminase IgA), sweat test, and genetic screens were each counted as 1 test.
At follow-up visits, anthropometric data, adherence to provider recommendations, and any new recommendations were recorded. A child was determined to be adherent only if adherent with all of the provider's recommendations as documented by the primary provider. Collected data was entered into REDCap (Research Electronic Data Capture) system database (14). Children were found to have nonorganic FTT when no organic cause was found or if a nonorganic etiology such as improper mixing of formula was discovered.
Descriptive statistics were used to describe the patient population. χ2 test for independence or Fisher exact test was used to describe categorical variable associations. Univariate analysis was conducted using Student t test or the Mann-Whitney test. Correlation of continuous variables was performed using Spearman rho. A P value of ≤0.05 was considered to be statistically significant for all tests and all of the tests were 2-sided. Because the majority of our study population was younger than 2 years of age, we also examined this group of children separately from the group as a whole. Statistical analysis was performed using SPSS statistical software system (version 21, IBM Corp, Armonk, NY). The institutional review board of Children's Hospital of Wisconsin approved the present study (IRBNet ID 462506-2).
During the 15-month study period, 110 patients were enrolled. Seventy (63.6%) were boys. The median age at consent was 0.79 years (IQR 0.36–1.98). Sixty-six (60%) of the patients were on public insurance. Sixty-two children (56.4%) were white, 30 (27.3%) were black, 6 (5.4%) were Asian, 4 (3.6%) were American Indian, and 3 (2.7%) belonged to other races. Child protective services were involved in 10 (9.1%) of the patients.
The most common FTT criterion met by enrolled children was weight declaration across >2 major percentiles since birth (n = 59) and the least common was weight <80% of median weight-for length (n = 17) (Table 1). The mean number of criteria met by enrolled children was 3.55 ± 1.67. Seven children met all of the 7 criteria and 16 children met only 1 criterion (Table 1). With regard to z score and caloric intake criteria for malnutrition, 74 patients (67.2%) were malnourished based on weight-for-length/BMI (z score ≤−1) and 14 patients (12.7%) were malnourished based on length/height (Table 2). Forty-six children (41.8%) had weight gain of <75% of expected for age and 64 (60.4%) had <75% of required caloric intake. Intake could not be assessed in 4 children because of them being exclusively breast-fed. In all, 92 children (83.6%) met anthropometric and intake and/or weight gain criteria for malnutrition. During the enrollment period, a total of 15 providers saw patients with FTT with the majority of patients (76) being seen by 1 of 2 providers in our nutrition clinic.
Forty-four patients (40%) had other major medical diagnoses at presentation to the pediatric GI clinic. Of these, 17 children had developmental delay. The remaining children with prior major diagnoses included pulmonary disease (n = 11), neurological disease besides developmental delay (n = 7), gastrointestinal disease (2 each with pyloric stenosis, severe constipation, and milk protein intolerance and 1 with gastroschisis), atopy (eczema or food and/or environmental allergies) (n = 7), attention deficit hyperactivity disorder (n = 5), laryngomalacia or upper airway abnormalities (n = 3), genetic/metabolic disorders (n = 2), hematological disease (n = 2), renal disease (n = 2), orthopedic diagnoses (n = 2), and congenital heart disease (n = 1).
Birth characteristics included 28 children (25.5%) who were small for gestational age and 5 children (4.5%) who were large for gestational age. Median gestational age was 39 weeks (IQR 38–40) with 19 children (17.3%) being premature. Six children (5.5%) were twins (1 set of twins and 4 children who had twin siblings without FTT).
Evaluation of FTT
When looking at all tests, 73 children (66.3%) had a total of 254 laboratory tests, imaging studies, and endoscopies performed that is equivalent to a median of 4 tests per patient (IQR 3–5). A total of 5 tests (6.8%) were positive. The median overall cost for tests was $1103 (IQR 568–3920).
Sixty-two patients (56.4%) had laboratory evaluation during the work-up of their FTT. A total of 200 laboratory tests were performed with a median of 3.5 tests per patient who had any laboratory tests done (IQR 2–4). Thirty-eight children had laboratory evaluation at their primary doctor's office (61.3%), 28 (45%) had laboratory tests drawn at the time of the pediatric GI clinic visit and 8 (12.9%) patients had laboratory tests drawn during a hospital admission. In 2 patients (3.2%), results from their laboratory evaluation led to an etiology of their FTT (chromosomal deletion and William syndrome). Both of these children were suspected of having a genetic etiology based on clinical examination. The median laboratory charges per patient (excluding cost of genetic testing) was $932.75 (IQR 542.23–1105.25).
Thirty-two children (29.6%) had a total of 40 imaging tests. The most common tests were an upper gastrointestinal barium study and chest x-ray (Table 3). The only positive imaging study was in a child who developed persistent vomiting after initially presenting with FTT and had a hypothalamic glioma found on magnetic resonance imaging (3.1%).
Twelve children (10.9%) underwent endoscopic evaluation (Table 3). Two of these children (16.7%) were diagnosed with eosinophilic esophagitis. In these children, the indication for endoscopy was concomitant vomiting and FTT. In addition to vomiting, one of the positive cases also had atopy. In children with normal endoscopy, the indication for the procedure in 2 of the children was FTT alone. For the other children with normal endoscopy results, 2 children had a history of feeding aversion, 4 children had a history of atopy with 1 of those also having history of vomiting, and 1 child had a history of diarrhea, all in addition to FTT. One child had endoscopic pancreatic stimulation testing and 3 children had disaccharidase analysis of duodenal biopsies performed. These additional tests were all normal.
The most common interventions recommended to patients were increasing calories, avoiding grazing, and structuring meals and snacks. Five children (4.5%) had enteral feeding tubes placed; all except 1 of these children were found to have organic etiologies of their FTT. These etiologies included significant feeding disorder associated with developmental delay (n = 1), chronic lung disease (n = 1), hypothalamic glioma (n = 1), and laryngomalacia (n = 1). The child with nonorganic failure to thrive had short-term nasogastric supplementation and once nutrition status improved, the child resumed full oral feeds.
Comparison of Organic and Nonorganic FTT
A final determination of the primary etiology of FTT was available for 91 patients; the other 19 patients did not have follow-up data available (Fig. 1). The 30 children lost to follow-up were not statistically different from the remaining study group in regards to demographic factors or anthropometric z scores at presentation (P > 0.05). Of the children with available follow-up, 81 (89%) had primary nonorganic etiologies for their FTT. Five of these children (5.5%) were felt to have FTT because of decreased appetite secondary to stimulant medications. The new organic etiologies of FTT were eosinophilic esophagitis or allergic colitis (n = 3), genetic abnormalities (n = 2), and hypothalamic glioma (n = 1). Four children were considered to have FTT because of a prior medical diagnosis, and these included chronic pulmonary disease (n = 2), galactosemia and vomiting (n = 1), and laryngomalacia (n = 1).
There was no significant difference in any of the demographic measures obtained between the organic and nonorganic FTT groups including age, sex, insurance type, the presence of a prior medical diagnosis, what FTT criteria each patient met, or the number of FTT criteria met. There was no difference in presenting anthropometric z scores for weight, length, weight-for-length, BMI, or OFC in those with organic and nonorganic FTT (Table 4). When looking at children younger than 2 years, there was also no difference in any of the above-mentioned factors. In addition, there was no difference in a child meeting the criteria for malnutrition or severity of malnutrition in children with organic and nonorganic FTT (P < 0.05).
Weight Gain in Children With Nonorganic FTT
Seventy children had nonorganic FTT and available follow-up information. Children with nonorganic FTT who were younger than 2 years had significantly shorter time to follow-up compared with those who were 2 years or older of 26 days (IQR 19–94) vs 128 days (IQR 66–152) (P < 0.001). Children who were adherent with provider recommendations gained significantly more weight than those who were nonadherent (Tables 5 and 6). Other factors were not associated with weight gain including sex, age, prematurity, presence of another medical diagnosis, insurance type, number of recommendations, or any of the individual recommendations (P > 0.05).
When looking specifically at the adherent and nonadherent groups, demographic factors including age, presence of a prior medical diagnosis, and insurance type were not significantly related to adherence with provider recommendations (Tables 5 and 6). Initial z scores for weight, height/length, weight for height, BMI and OFC and FTT criteria, and number of criteria met for those patients who were adherent and those who were not were also not statistically different between the 2 groups (Tables 5 and 6). In addition, the number of individual provider recommendations (Tables 5 and 6) or specific recommendations given was not significantly different between the adherent and nonadherent group (P > 0.05). There was no difference in median time to follow-up for those who were adherent compared with those who were nonadherent (Tables 5 and 6).
Our study confirms that the majority of English-speaking patients referred for FTT to a tertiary pediatric gastroenterology clinic, like children in other settings, have nonorganic causes for their FTT (3–5,9,15). Our data suggest that in most cases, laboratory evaluation does not change the diagnosis or management of children with FTT. We also have shown that standard behavioral interventions are associated with weight gain when families are adherent with these recommendations.
Previous studies of inpatients have found that 13.9 to 39.6 tests per patient were performed in the evaluation of FTT with 0.8% to 1.4% of these tests aiding in the diagnosis of FTT (4,5). In our population, more than half of the children had tests performed in evaluating their FTT (66.3%) but the number of tests per patient in our study population was fewer than previously reported at 4 (IQR 3–5) per patient who had any study performed. This may be due to the fact that our study evaluated outpatients, whereas the prior studies looked at inpatient and likely sicker populations. It also may reflect an increased awareness of the limited utility of testing or may be due to different ways of counting individual tests. In our cohort, only 2 children (3.2%) had positive laboratory tests that led to an etiology of FTT. Both were genetic evaluations initiated based on developmental history and physical examination.
Despite an increasing prevalence and awareness of celiac disease (16), none of our children with FTT were found to have celiac disease. As most of our patients with FTT and laboratory evaluation had these studies done by their primary care providers, we speculate that patients with celiac disease may have been referred to the pediatric GI clinic for a positive celiac screen and not necessarily for FTT. The lack celiac disease among our patients may also reflect a trend in the earlier detection of celiac disease before the onset of poor growth (16).
In our cohort, 1 child had a positive head magnetic resonance imaging out of a total of 40 imaging tests performed (2.5%). Similarly Sills et al (4) reported that 4.4% of imaging studies performed led to a diagnostic etiology of FTT in their patients although these results were almost always suspected based on history and examination. Although Berwick et al (5) reported more positive tests related to GI imaging, we did not have any positive GI imaging results.
Endoscopic evaluation of children with FTT has not been well described. In our study, the utility of endoscopy was higher than that of imaging or laboratory evaluation (16.7% diagnostic for organic disease); however, endoscopy was the least used investigative modality being used in only 10% of our patients. The most common reason for performing endoscopy in our cohort of children with FTT was the additional symptom of vomiting due to the concern for eosinophilic esophagitis. None of the children who underwent endoscopy for FTT without associated symptoms had positive endoscopy results. The higher diagnostic value of endoscopy compared with imaging and laboratory evaluation may represent more judicious use of this invasive testing modality. More research needs to be done assessing the utility of endoscopy in children with FTT.
Overall evaluation charges amounted to just over $1103 per patient. This is an underestimate of the total charges because we excluded the charges generated by expensive genetic testing and physician charges. Our study cannot be used to determine the optimum workup of children with FTT. We can only state that in most cases testing adds to the cost of patient care without changing the diagnosis or management of children with FTT. Our data suggest that judicious testing of patients based on history and physical examination is important. We also think it is important to follow children with FTT to ensure that growth improves. Testing may be warranted if new symptoms develop and/or FTT does not resolve especially in a family that is adherent with provider recommendations.
Our patients are clearly different from children described in previous studies because the present study was undertaken in a tertiary care center pediatric GI clinic. Almost 40% of the patients referred to our clinic had a significant prior medical diagnosis. Even among these children, only 4 children of 44 (9.1%) had FTT secondary to that prior medical diagnosis. The remaining 40 children were felt to have primary nonorganic reasons for their poor growth despite having a significant prior medical diagnosis. The low rate of organic FTT in children with prior medical diagnoses may be due to the fact that if the poor growth is the result of another disease, the underlying disease may be treated by the primary physician or specialist without referral to a gastroenterologist. These results highlight the importance of considering nonorganic etiologies for poor growth even in children with complex disorders involving other organ systems.
The medical management of FTT in a pediatric GI clinic has not been previously described. Many of the children with nonorganic FTT likely have restrictive or other feeding disorders (17). We found that adherence to provider recommendations was significantly associated with a positive change in weight z score but each individual recommendation did not predict growth. We postulate that the lack of statistical significance for individual interventions is due to the fact that the interventions are not used in isolation and there are likely other factors that influence growth in these children. In addition to behavioral interventions in our clinic, calorie provision is increased by concentrating the formula in infants, providing 30 cal/oz (1 kcal/mL) beverages to toddlers and older children, and adding fats to solid foods. Multivitamin use was occasionally suggested to assure complete nutrition, and cyproheptadine was used in a subset of patients as an appetite stimulant (18).
A major issue with any medical intervention is adherence to provider recommendations. In our study, adherence was the only factor significantly associated with weight gain. One study showed similar results in premature infants with FTT, those who were adherent to a multifaceted intervention had better growth outcomes at 3 years of age compared with those that were not adherent or were in the nonintervention group (19). This emphasizes the importance of adherence and may help to explain why home-health visits and a multidisciplinary approach have been shown to be beneficial in reversing FTT (9,20,21).
The majority of the children in the present study were younger than 2 years. When looking at the younger children, the only difference compared with older children was that their time to follow-up was shorter. This is expected because of the more rapid growth seen in younger children. There were more older children with nonorganic FTT secondary to stimulant medications (24% vs 0%); this is most likely due to the fact that these medications are almost exclusively used in older children for treatment of attention-deficit disorders. Our data suggest that primary nonorganic causes of FTT predominate in both groups.
This is a referral population of patients, so the generalizability of the results to a general pediatric population is unknown. In addition, only English-speaking patients were included in the present study, further limiting the generalizability of the study results. Limited socioeconomic data were collected in the present study and further research into socioeconomic influencers in FTT may help define this patient population. The present study was not designed to determine which children with FTT need further evaluation. Despite this, we feel it may be prudent to consider further evaluation in children with other symptoms and/or if a child does not show weight gain despite adherence to behavioral interventions. Given the descriptive nature of the present study, we are not able to validate specific interventions but have been able to show improved growth with implementation of provider recommendations. Lastly, we do not know if these growth outcomes are sustained over the long term because data were examined at the first follow-up visit. Despite these weaknesses, the present study provides a snapshot of FTT in the pediatric GI clinic.
The present study shows that the majority of cases with FTT referred to a tertiary pediatric gastroenterology clinic have nonorganic etiologies of their FTT. It also confirms that laboratory, imaging, and endoscopic evaluations rarely change the diagnosis or management of children with FTT unless there is clinical suspicion based on history and physical examination. Also, a standardized set of behavioral and nutritional interventions likely positively impacts weight gain in children with FTT. Despite technological and diagnostic advances during the past 30 years, it still holds true that less is more when evaluating a child with FTT.
1. Jaffe AC. Failure to thrive
: current clinical concepts. Pediatr Rev
2. Olsen EM, Petersen J, Skovgaard AM, et al Failure to thrive
: the prevalence and concurrence of anthropometric criteria in a general infant
population. Arch Dis Child
3. Mitchell WG, Gorrell RW, Greenberg RA. Failure-to-thrive: a study in a primary care setting. Epidemiology and follow-up. Pediatrics
4. Sills RH. Failure to thrive
. The role of clinical and laboratory evaluation. Am J Dis Child
5. Berwick DM, Levy JC, Kleinerman R. Failure to thrive
: diagnostic yield of hospitalisation. Arch Dis Child
6. Fryer GE. The efficacy of hospitalization of nonorganic failure-to-thrive children: a meta-analysis. Child Abuse Negl
7. Blair PS, Drewett RF, Emmett PM, et al Family, socioeconomic and prenatal factors associated with failure to thrive
in the Avon Longitudinal Study of Parents and Children (ALSPAC). Int J Epidemiol
8. Daniel M, Kleis L, Cemeroglu AP. Etiology of failure to thrive
in infants and toddlers referred to a pediatric endocrinology outpatient clinic. Clin Pediatr
9. Atalay A, McCord M. Characteristics of failure to thrive
in a referral population: implications for treatment. Clin Pediatr
11. Becker P, Carney LN, Corkins MR, et al Consensus statement of the Academy of Nutrition and Dietetics/American Society for Parenteral and Enteral Nutrition: indicators recommended for the identification and documentation of pediatric malnutrition (undernutrition). Nutr Clin Pract
12. Bocca-Tjeertes IFA, Reijneveld SA, Kerstjens JM, et al Growth in small-for-gestational-age preterm-born children from 0 to 4 years: the role of both prematurity and SGA status. Neonatology
13. WHO Anthro for personal computers [computer program]. Version 3.2.2. Geneva: World Health Organization; 2011.
14. Harris PA, Taylor R, Thielke R, et al Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform
15. Shaheen E, Alexander D, Truskowsky M, et al Failure to thrive
—a retrospective profile. Clin Pediatr
16. Murray JA, Van Dyke C, Plevak MF, et al Trends in the identification and clinical features of celiac disease in a North American community, 1950–2001. Clin Gastroenterol Hepatol
17. Kerzner B, Milano K, MacLean WC, et al A practical approach to classifying and managing feeding difficulties. Pediatrics
18. Couluris M, Mayer JLR, Freyer DR, et al The effect of cyproheptadine hydrochloride (periactin) and megestrol acetate (megace) on weight in children with cancer/treatment-related cachexia. J Pediatr Hematol Oncol
19. Casey PH, Kelleher KJ, Bradley RH, et al A multifaceted intervention for infants with failure to thrive
. A prospective study. Arch Pediatr Adolesc Med
20. Black MM, Dubowitz H, Hutcheson J, et al A randomized clinical trial of home intervention for children with failure to thrive
21. Black MM, Dubowitz H, Krishnakumar A, et al Early intervention and recovery among children with failure to thrive
: follow-up at age 8. Pediatrics
Keywords:© 2016 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,
child; endoscopy; failure to thrive; infant