The severity of malnutrition in children with epidermolysis bullosa correlates with disease severity

WHO defines malnutrition as severe if the z-scores are less than − 3 Standard deviation (SD), moderate if between − 2 and − 3 SD and mild if between − 2 SD to – 1 SD. This study was aimed to assess nutritional aspects of Indian children suffering from EB and to evaluate the effect of severity of EB on the severity of malnutrition. In this study, pediatric EB patients were evaluated prospectively for baseline nutritional status using anthropometric parameters and WHO growth charts, and its correlation with disease severity using instrument for Scoring Clinical Outcomes for Research of Epidermolysis Bullosa-iscorEB. In second phase, an individualized diet chart was given to meet the energy, protein and micronutrients needs and its effects were observed after 6 months. The median age of participants was 3 years (IQR-9). Of 57 patients, malnutrition was seen in 40.35% patients (22.81%-moderate and 17.54%-severe), and significantly correlated with iscorEB (r = 0.45, p < 0.0001). On bivariate regression analysis, iscorEB was independently associated with moderate-to-severe malnutrition (p = 0.047; OR 1.038, CI 1.011–1.066). iscorEB enabled the identification of patients with moderate-to-severe malnutrition with an Area Under Receiver Operating Curve (AUROC) of 0.72 (95%CI 0.58–0.85; p < 0.005). In phase 2, there was significant improvement in nutritional status in children with recessive dystrophic EB (RDEB) and dominant dystrophic EB (DDEB) subtype (p < 0.0001). The severity of malnutrition in EB children significantly correlates with disease severity, and is an independent predictor of moderate-to-severe malnutrition.

Phase I: Anthropometry and disease severity. A detailed history followed by a complete physical examination was recorded in case record proforma. The severity of EB was calculated using 'iscorEB' , including clinician and patient score 6 . Besides, physical examination was focused on the anthropometric examination like weight, height, head circumference, mid-arm circumference, and weight for age, height for age, weight for height and body mass index (BMI) and comparison with the age-appropriate World Health Organization (WHO) growth charts to assess whether the child had any features of malnutrition 7 . They were categorized into having severe malnutrition if the z-score (weight for height and height for age) are less than -3 Standard deviation (SD), moderate if z-score are between -2 and -3 SD, mild if values are between -2 SD to -1SD 1 . Table 1 enlists the laboratory investigations done in patients.
Phase II: Nutritional advice. The objective of the nutritional advice for children with Epidermolysis Bullosa (EB) was to provide more than adequate macronutrients as well as micronutrients, minimize the nutritional deficiencies, improve bowel functions, enhance immunological status, optimize wound healing, promote proper body growth as well as pubertal and sexual development and also is to minimize the stress caused by prolonged feeding time.
Prior to nutritional advice given to parents of an EB child, a detailed history was documented in a precise format with special emphasis on dietary intake with the help of 24-h recall method. Dietary history included typical meal pattern, consistency of food, frequency of meals, time taken over average meals, nutrient supplements if any, reason for difficulty in food intake according to child/ caretaker. Nutrient value of dietary intake and requirements were assessed as per guidelines by National Institute of Nutrition, Hyderabad 8 . After this, an individualized diet chart was given to meet the energy, protein and other nutrients needs of the child with EB and www.nature.com/scientificreports/ also more emphasis were given for frequent follow ups to monitor the improvement of the child (supplemental Fig. 1). Children with EB often requires more calories than usual. Thus, their diet plan was made by providing 100% to 150% of the Estimated Energy Requirement (EER) which is part of Dietary Reference Intake (DRI). If growth was not satisfactory based on follow up assessment, a gradual increase in energy was initiated. Due to excess protein loss through blisters as well as recurrent infections, the protein needs of the child with EB are greater than those healthy peers of matching gender and age. To maintain positive nitrogen balance, diet plans were calculated by providing 115-200% of the estimated average requirement as per Recommended Dietary Allowance (RDA) 9 . Adequate amounts of carbohydrate, especially complex carbohydrates were included in their daily diet plan to overcome the problem of constipation. Also, adequate amounts of fats such as butter/ghee if needed were provided as per RDA in their menu to increase the calorie need and palatability. While planning menu foods rich in vitamins like A, C, D, B6, B12, thiamine, riboflavin and minerals like Zinc, calcium and iron were well taken care in their daily diet.
At the end of the diet counseling, parents of child with EB were imparted the knowledge of high energy giving foods and how to prepare delicious and nutritious recipes in different consistency (thick liquid, semisolid, pureed) to improve their child's taste and intake and given recipe handouts written in Hindi/English/Punjabi.
Since frequency, quantity and consistency of meals/feeds are very important factors while giving nutritional advice to children with EB, parents were advised to give small frequent meals/feeds customized to local tastes. In   10 , daily supplementation of a common commercially available multivitamin syrup was also supplemented. Micronutrient deficiency detected in laboratory evaluation was treated as per pediatrician advice. Esophageal dilation was done in patients with esophageal strictures detected on barium swallow by a pediatric gastroenterologist. Patients were followed up every month for at least 6 months. Repeat anthropometry and laboratory investigations were done at 6 months.
Statistical analysis. The normality of quantitative data was checked by measures of Kolmogorov Smirnov tests of normality. Normally distributed data were compared using one-way ANOVA followed by post hoc multiple comparisons test. For the skewed data such as the baseline clinical and anthropometric variables, the Kruskal-Wallis test followed by the Mann-Whitney test was applied ( Table 2). For comparison of time-related variables of skewed data such as the serial change in anthropometric characteristics, the Wilcoxon signed-rank test was carried out (Table 1S). Spearman correlation coefficients were calculated to see the relationship between different variables. To find an independent predictor for moderate to severe malnutrition, logistic regression analysis was carried out. Area Under Receiver Operating Characteristic (AUROC) curve was calculated to find maximal cut-off values of iscorEB to detect malnutrition. All statistical tests were two-sided and performed at
We tried to find out factors associated with moderate to severe malnutrition. On bivariate analysis, it was significantly associated with gastrointestinal system involvement (p = 0.032) and iscorEB (p = 0.005) but not with age (p = 0.15), gender (p = 0.71), disease subtype (p = 0.12), and dental involvement (p = 0.07). On bivariate regression analysis, only iscorEB was independently associated with moderate to severe malnutrition (p = 0.047; OR = 1.038, CI = 1.011-1.066). iscorEB enabled the identification of patients with moderate to severe malnutrition with an AUROC of 0.72 (95% CI 0.58-0.85; p < 0.005, Fig. 3). At an estimated cut-off value of 24.5, iscorEB exhibited a sensitivity and specificity of 82.6% and 55.9%. Using the logistic regression model, an iscorEB value of more than 24.5 was associated with a markedly increased risk of moderate to severe malnutrition (p = 0.05; OR = 5.12, 95% CI = 1.00-26. 15). Forty-one out of 57 patients underwent laboratory evaluation (Table 3). Eight patients had mild anemia (10-11 g/dl), 18 patients had moderate anemia (7-10 g/dl) and 1 patient had severe anemia (< 7 g/dl). Vitamin B12 deficiency was observed in 20 patients. Mean iscorEB was higher in those with moderate anemia (48.71 ± 24.20) and severe anemia (50.00) compared to those with mild anemia (23.36 ± 20.28)   15 and 20). Of the 57 patients, 34 who followed dietary advice as advised by a dietician experienced a significant increase in mean weight and height (P < 0.0001). Mean follow-up duration was 6 months. Mean weight increased from 15.53 ± 15.02 kg to 17.87 ± 14.71 kg, and mean height increased from 92.64 ± 36.94 cm to 97.55 ± 34.14 cm. Similarly, baseline weight for age increased significantly from − 1.74 ± 1.13 to − 1.33 ± 1.10 at the end of 6 months with a corresponding increase in Z-scores (P = 0.01). Height for age increased from − 1.47 ± 0.98 to − 1.28 ± 1.11; and weight for height increased from − 1.43 ± 1.19 to − 1.04 ± 0.97 but improvements in these parameters were statistically non-significant (P = 0.11 and P = 0.10 respectively; Wilcoxan-signed rank test) (Table 1S). Among these 34 patients, the frequency of severe malnutrition decreased significantly from 26.4% to 8.8%. Improvement in nutritional status was statistically significant in the overall study cohort particularly due to significant improvement in RDEB (p < 0.0001) and DDEB (p < 0.009) subtypes (supplementary table 1).

Discussion
The nutrition in the first few years of the life of a child is crucial concerning the overall growth and development and any factor that interferes with adequate nutritional intake during this period predisposes to significant failure to thrive. In the present study, malnutrition was seen in 40.35% patients, and significantly correlated with severity of disease. In fact, the disease severity was an independent determinant of severity of malnutrition. Also, there was significant improvement in nutritional status in children with dystrophic EB (RDEB) and dominant dystrophic EB (DDEB) with dietary advice. In our study, the median age at presentation of 3 years though comparable to other Indian studies 11 , reflects the delay in presentation to a tertiary care institute with specialized EB services in a developing country due to poor access to health care, and a lack of information and knowledge about the disease among parents and primary care physicians. There was a male predominance in the present study (M:F = 1.3:1) comparable to another Indian study by Hiremagalore et al. 12,13 .
Pediatric malnutrition can be illness-related and non-illness-related. Since there was no control group in the present study, we compared our data with the data in the National Family Health Survey 4(NFHS 4) 13 , which showed that 35.7% of Indian children below 5 years were underweight, 38.4% were stunted, 21% were wasted. Compared to children in NFHS 4 survey, a higher frequency of EB children was underweight and wasted with a slightly higher frequency of stunting. This can be attributed to a wide range of gastrointestinal complications affecting EB children in 69.2% of patients with moderate malnutrition and 90% of patients with severe malnutrition. Dental involvement was seen in a much higher proportion of patients with moderate to severe malnutrition although it did not achieve statistical significance. In EB patients, reduced food intake and enhanced nutritional needs together contribute to significantly enhanced risk of malnutrition, leading to interruption in growth, delayed puberty and wound healing 3,9 . Hence, the management of such comorbidities can help improve the nutritional status of EB children. Severe malnutrition was seen predominantly in females (M:F ratio = 2:3). This female predominance in malnutrition is multifactorial and is also reported in NFHS survey 14 .
Severe malnutrition was most commonly seen in RDEB (26.9%), followed by DDEB (25%), and JEB (12.5%). However, the disease subtype was not found to have a significant bearing on the frequency of malnutrition (p = 0.10). Few previous studies on this aspect have also reached similar conclusions with the frequency of malnutrition and failure to thrive being higher in patients with recessive dystrophic EB and junctional. Birge et al. found that out of 80 EB children, only 22% with EBS were having malnutrition whereas 77% with dystrophic EB and 57% with junctional EB had malnutrition. Fine et al. reported that 39% of RDEB, 2% of DDEB, 78% of children with JEB Herlitz type and 43% of those with non-Herlitz type were found to have malnutrition [15][16][17] . On correlating malnutrition with iscorEB, there was a statistically significant correlation with a moderate strength of association. Additionally, iscorEB was found to have an independent association with malnutrition (p = 0.005; OR = 1.038, 95%CI = 1.011-1.066) with a much stronger association at a cut-off level of 24.5. Since the disease severity goes hand in hand with malnutrition risk, it also alerts the treating physician about the need for an urgent nutritionist or pediatric gastroenterologist consultation.
The mean iscorEB was found to be higher in those with moderate and severe anemia compared to those with mild anemia. According to the Australasian EB registry 18 , 25.8% of pediatric EB patients were anemic and mostly seen in RDEB and JEB generalized severe type. Nearly 30% of patients had vitamin D deficiency. Mean iscorEB was higher in those with deficiency followed by those with insufficient levels (45.12 ± 27.61) compared to those with sufficient vitamin D levels (26.20 ± 22.97) but was not statistically significant (P < 0.259). Hypocalcemia could be due to decreased dietary intake or decreased absorption and thus affecting bone mineralization. The cause for hypermagnesemia could not be identified and patients did not manifest any symptoms. These observations are similar to other studies [19][20][21][22][23] .
To ensure supplementation of required calories, proteins and fibers, consistency, frequency, type of food were individualized based on age, severity of EB, severity of dysphagia. The total amount of calories was slowly built up by gradually increasing every week to reach the required level. This helped to overcome difficulties in dietary intake in children. For those with dysphagia, esophageal dilation was done which helped to increase the dietary intake to required level.
After following dietary advice for 6 months, there was a significant improvement in weight and height compared to baseline. After following dietary advice for 6 months, 29.4% had malnutrition ( www.nature.com/scientificreports/ found improvement in weight for age, height for age and weight for height scores after following dietary advice for 9-18 months when compared with the control group 24 . Since our follow up duration was only 6 months, an extended follow-up period might show even greater improvement in stunting and other anthropometric parameters. The improvement in nutrition especially in those with malnutrition by offering 100% to 150% of the EER and 115-200% of the estimated average requirement of protein underscores the increased demands to be met in EB children compared to their normal counterparts. Improvement in the frequency of malnutrition following dietary advice highlights the importance of diet in the management of overall growth and severity of disease in EB children. Better dietary intake helps to counter the detrimental effects of disease complications and increased demands because of the chronic inflammatory state. The barriers towards attaining normal nutritional status in 29.4% of patients in spite of dietary management could be the complications associated with the disease, inadequate parental education, and poor socioeconomic status and lack of initiative and encouragement at home for the same. Despite our best efforts, our study had a few limitations. The results should be validated in a much larger group in the developing world preferably in a multicentric setting. Moreover, the effect of dietary advice was not compared with the control group as the children normally tend to gain weight and height during this period of growth and development.