Murdoch Childrens Research Institute, Parkville, Vic., Australia.
Department of Paediatrics, University of Melbourne, Parkville, Vic., Australia.
Clin Exp Allergy. 2015 May;45(5):953-963. doi: 10.1111/cea.12478.
Food allergy, eczema and wheeze are early manifestations of allergic disease and commonly co-occur in infancy although their interrelationship is not well understood. Data from population studies are essential to determine whether there are differential drivers of multi-allergy phenotypes. We aimed to define phenotypes and risk factors of allergic disease using latent class analysis (LCA).
The HealthNuts study is a prospective, population-based cohort of 5276 12-month-old infants in Melbourne, Australia. LCA was performed using the following baseline data collected at age 12 months: food sensitization (skin prick test ≥ 2 mm) and allergy (oral food challenge) to egg, peanut and sesame; early (< 4 months) and late-onset eczema; and wheeze in the first year of life. Risk factors were modelled using multinomial logistic regression.
Five distinct phenotypes were identified: no allergic disease (70%), non-food-sensitized eczema (16%), single egg allergy (9%), multiple food allergies (predominantly peanut) (3%) and multiple food allergies (predominantly egg) (2%). Compared to the baseline group of no allergic disease, shared risk factors for all allergic phenotypes were parents born overseas (particularly Asia), delayed introduction of egg, male gender (except for single egg allergy) and family history of allergic disease, whilst exposure to pet dogs was protective for all phenotypes. Other factors including filaggrin mutations, vitamin D and the presence of older siblings differed by phenotype.
Multiple outcomes in infancy can be used to determine five distinct allergy phenotypes at the population level, which have both shared and separate risk factors suggesting differential mechanisms of disease.
食物过敏、湿疹和喘息是过敏疾病的早期表现，在婴儿期常同时发生，尽管它们之间的关系尚不清楚。人群研究的数据对于确定多过敏表型的不同驱动因素至关重要。我们旨在使用潜在类别分析（LCA）定义过敏疾病的表型和危险因素。
HealthNuts 研究是一项前瞻性、基于人群的队列研究，纳入了澳大利亚墨尔本 5276 名 12 个月大的婴儿。在 12 个月大时收集以下基线数据进行 LCA：对鸡蛋、花生和芝麻的食物过敏（皮肤点刺试验≥2 毫米）和过敏（口服食物挑战）；早期（<4 个月）和晚期发作的湿疹；以及婴儿期第一年的喘息。使用多项逻辑回归模型来分析危险因素。
确定了五种不同的表型：无过敏疾病（70%）、非食物过敏的湿疹（16%）、单一鸡蛋过敏（9%）、多种食物过敏（主要是花生）（3%）和多种食物过敏（主要是鸡蛋）（2%）。与无过敏疾病的基线组相比，所有过敏表型的共同危险因素包括父母出生在海外（尤其是亚洲）、鸡蛋延迟引入、男性（除了单一鸡蛋过敏）和家族过敏史，而接触宠物狗对所有表型都有保护作用。其他因素，包括丝聚蛋白突变、维生素 D 和有年龄较大的兄弟姐妹，因表型而异。
婴儿期的多种结局可用于在人群水平上确定五种不同的过敏表型，这些表型具有共同和单独的危险因素，表明疾病的发病机制不同。
