Poor head growth and developmental delay in infants with eczema, food allergies and growth faltering

To the Editor, Earlyonset, moderatetosevere eczema in infants is associated with food allergy in 30– 65% and faltering growth in infancy.1 We observed poor head growth and developmental delay in severely affected infants and conducted a retrospective case review to compare clinical and laboratory features between infants with severe eczema, food allergy/sensitisation and faltering growth with restricted head growth (cases) and those with similar clinical presentation but preserved head growth (comparator/control group). A clinical database search (Paediatric Allergy Service, St Mary's Hospital, London UK) for the period 2002– 2011 was undertaken with the search terms: ‘growthfaltering/failure’, ‘poor weight gain’, ‘failure to thrive’, ‘malnutrition’ and ‘stunting’. The hospital review and audit board approved the study. Inclusion criteria were age < 18 months at presentation; severe eczema, faltering growth, food allergy/sensitisation. Exclusion criteria were insufficient data available to determine infants' growth pattern. Records were examined independently by two paediatricians (EM/CG). Neurodevelopmental delay was diagnosed if neurological examination by >1 certified/senior paediatrician showed that ageappropriate developmental milestones were not achieved. Eczema was diagnosed and severity determined by the degree of skin involvement, sleep disturbance, need for admission and potent topical corticosteroids. Food allergy was diagnosed based on a history of rapid reaction on exposure to food with corroborating positive specific IgE (ImmunoCAP; Phadia, Uppsala/Sweden; ≥0.35 kUA/L considered positive) or skin prick tests (Stallergenes, France; wheal diameter of ≥3 mm above saline control considered positive) and/ or tests exceeding previously published 95% threshold values for reactivity. Testpositive foods which failed to fulfil the above criteria were initially avoided and subsequently introduced by supervised oral food challenge. Elimination diets were supervised by experienced paediatric dietitians. Weight, height and occipitalfrontal head circumference (OFC) were measured according to accepted standards and documented on sexspecific United Kingdom Growth Charts. Dietetic and personal child health records yielded additional measurements. Weight and OFC measurements were converted into zscores using WHO Anthro (3.2 WHO, 2006), and definitions for moderate malnutrition of <−2 zscores were used.2 Growth faltering was defined as a change in weightforage of > − 1 zscore in a 3month period in infants less than 1 year of age. In children over 1 year of age, growth faltering was defined as inadequate growth or weight gain ≥1 month in a child under 2.2 Moderate nutritional deficit was defined as >−2 zscores using World Health Organisation criteria.3 Nutrition rehabilitation was considered to have been achieved when weightforage and OFCforage were >−1 zscore. Other study terminology included prediagnosis zscores, highest zscore recorded prior to clinical presentation; diagnosis (acute) zscores, taken at the time of clinical presentation and allergy diagnosis; convalescent zscores, taken 4– 60 weeks after initiation of allergenexclusion measures. Cases were defined as infants with ≥0.67 standard deviations decrease in OFCforage zscore and/or increase in OFCforage zscore following the institution of a hypoallergenic diet. Controls were defined as infants with a decrease of ≥ − 1 weightforage zscore over a period of 3 months and maintained OFCforage zscores. The chisquare test was used to compare categorical variables and nonparametric analysis for continuous variables (SPSS version 15.0; SPSS Inc, Chicago, IL, USA). Interpretation of analyses took account of the number of comparisons made, but formal correction for multiple comparisons was not used. We identified 65 infants aged < 18 months with documented faltering growth, severe eczema and multiple food allergy/sensitisation. OFC was recorded for 43/65 (66.2%), and sufficient growth data were available for 16/43 (37.2%). 15/16 infants presented to the emergency department from the community; one to the allergy service. All infants exhibited faltering growth with a severe decrease in weightforage at presentation (median zscore −2.13 [IQR −2.74 to −1.22]; range −1.01 to −4.62). Nine infants (56%) had restriction of head growth (cases) with decreased OFCforage (median zscore − 1.12 [IQR −1.29 to −0.63]; Figure 1A,B). There was no difference between cases and controls for demographics, family history of atopy, feeding practice, age of eczema onset, age at presentation or disease duration or weightforage zscores at diagnosis (Table 1). Cases exhibited a greater decrease in weightforage zscore between prediagnosis and diagnosis (median − 2.46 [IQR −2.97 to −1.67], range − 4.62 to −1.22) compared with controls (median − 1.28 [IQR −2.24 to −1.08], range − 2.4 to −1.01); p = .037. After treatment


Poor head growth and developmental delay in infants with eczema, food allergies and growth faltering
To the Editor, Early-onset, moderate-to-severe eczema in infants is associated with food allergy in 30-65% and faltering growth in infancy. 1 We observed poor head growth and developmental delay in severely affected infants and conducted a retrospective case review to compare clinical and laboratory features between infants with severe eczema, food allergy/sensitisation and faltering growth with restricted head growth (cases) and those with similar clinical presentation but preserved head growth (comparator/control group).  Growth faltering was defined as a change in weight-for-age of > − 1 z-score in a 3-month period in infants less than 1 year of age. In children over 1 year of age, growth faltering was defined as inadequate growth or weight gain ≥1 month in a child under 2. 2 Moderate nutritional deficit was defined as >−2 z-scores using World Health Organisation criteria. 3 Nutrition rehabilitation was considered to have been achieved when weight-for-age and OFC-for-age were >−1 z-score. Other study terminology included pre-diagnosis zscores, highest z-score recorded prior to clinical presentation; diagnosis (acute) z-scores, taken at the time of clinical presentation and allergy diagnosis; convalescent z-scores, taken 4-60 weeks after initiation of allergen-exclusion measures. Cases were defined as infants with ≥0.67 standard deviations decrease in OFC-for-age zscore and/or increase in OFC-for-age z-score following the institution of a hypoallergenic diet. Controls were defined as infants with a decrease of ≥ − 1 weight-for-age z-score over a period of 3 months and maintained OFC-for-age z-scores. The chi-square test was  Figure 1A,B). There was no difference between cases and controls for demographics, family history of atopy, feeding practice, age of eczema onset, age at presentation or disease duration or weight-for-age z-scores at diagnosis (Table 1).
Cases exhibited a greater decrease in weight-for-age z-score between pre-diagnosis and diagnosis ( Introduction of dietary and skin treatment led to the control of eczema, gastrointestinal symptoms and rapid catch-up growth in all infants (Figure 1). Inflammatory parameters returned to normal, with only mild eosinophilia persisting in most infants. Cases had a significantly higher z-score increase in convalescent OFC-for-age compared with controls and improvement in weight-for-age z-scores on treatment was higher for cases than controls ( Figure 1C). Infants with feeding aversion (n = 4 cases, n = 4 controls) had lower weightfor-age z-score increases post-treatment. Neurodevelopmental sequelae persisted for 3/5 children for ≥24 months.
Severe eczema and multiple food allergies are known to cause growth faltering in infancy and can be associated with allergic gastrointestinal inflammation. 1,4 The assumption hitherto has been that head growth is maintained as occurs with other malnutrition states. To our knowledge, this is the first description of restricted head growth with associated developmental delay in young infants presenting with severe, multisystem allergic disease, raised inflammatory parameters and multiple food allergy/sensitisations whilst being exclusively or predominantly breastfed.
Poorer growth in children with food allergies and eczema is recognised, 4 and negative impact on weight, height, body mass index and head circumference, but not neurodevelopmental progress, has been described. 5 Allergic inflammation may contribute to growth failure 4 and it has been linked with lower neuropsychological function, albeit within the normal range of development. 6 Other chronic inflammatory conditions such as juvenile arthritis, inflammatory bowel disease and Crohn's disease can lead to growth faltering in children with decreased weight gain and height velocity but not with cerebral effects and head growth tends to be preserved. 7 Cessation of head growth has been described in young children with chronic renal failure who displayed some additional developmental delay. 8 This was attributed to nutritional and inflammatory aberrations and appeared to improve post-transplantation.
Long-term follow-up is needed, as growth faltering at 6-8 weeks of life is associated with developmental delay and poor growth can impact negatively on cognitive ability and socio-economic activity in adulthood. 9 We hypothesise that systemic allergic inflammation may have profound effects on the brain in some infants and may lead to reduced brain growth until the inflammation is controlled.
This retrospective case review relied on data obtained at clinical presentation and from records, without formal eczema or growth assessment at predefined time points. We only included infants for whom extensive clinical data were available and defined growth faltering and restricted head growth according to accepted criteria. 3 In this retrospective study, we could not exclude specific genetic syndromes; however, neither clinical features nor progress of our cohort were typical of these.
Multiple statistical testing was performed to identify whether there were differences in sensitisation or overall inflammatory patterns. We interpreted the findings with caution but believe them to be important as the basis of future hypothesis development and study of underlying mechanisms. TA B L E 1 Differences between cases and controls for demographic details, feeding history, clinical presentation and treatment.

Cases n = 9
Controls n = 7 p-value a