The link between psychosocial stress and eczema (aka atopic dermatitis) was identified at least a half-century ago when eczema was designated one the seven classic psychosomatic disorders.[1] The allostatic model of chronic stress proposes that dysregulation of normal homeostatic mechanisms may lead to chronic hyperarousal or hyporesponsiveness.[2] Mechanistic evidence comes from the field of psychoneuroimmunology linking the nervous system, endocrine system, and immune system.[3,4] Stress may also lead to disease-exacerbating behaviors related to diet, sleep, exercise, bathing practices, and treatment adherence.[5] In epidemiologic studies, self-reported stress levels have been shown to correlate with eczema incidence, and stress-reducing modalities have been found to improve eczema symptoms.[6,7]

Various factors are involved in a stress response and will be included in our analyses: exposure to a stressor, resilience/ the use of social supports, and the emotional and behavioral response.[2] We will also include family-related distress such as caregiver stress, maternal panic disorder or depression and family conflict because these have been reported to have immunological effects on younger subjects.[8]

Aim: To examine the association between stress and eczema.

Hypothesis: Higher rates of distressing life events, poor social supports, and psychological distress will each be associated with more persistent eczema.

All patients in the ALSPAC cohort with follow-up data available through age 18 will be included in a longitudinal cohort study. Because the relationship between stress and eczema is likely bidirectional,8 we will model the relationship between each measure of stress and repeated measures of eczema persistence using a marginal structural model and stabilized inverse probability-of-treatment weights to avoid conditioning on stress through its inclusion as a predictor in the outcome model.[9] We will have 90% power to detect an odds ratio as small as 1.23 at a 0.05 significance level in a design with 5 repeated measurements, assuming 15% eczema prevalence.[10,11] We will carefully evaluate missingness and will conduct sensitivity analyses using multiple imputations.[12,13] The results of this study promise to disentangle the individual experience of stress from group-level factors that may be associated with stress including race, socioeconomic status, and environmental factors.