Background
Research has linked homelessness with an increased risk of skin conditions. However, representative studies of diagnosis-specific information on skin conditions in people experiencing homelessness are lacking.

Objectives
To examine the association between homelessness and diagnosed skin conditions, prescribed medication and type of consultation.

Methods
This cohort study included data from the Danish nationwide health, social and administrative registers from 1 January 1999 to 31 December 2018. All people of Danish origin living in Denmark and aged at least 15 years at some point during the study period were included. Homelessness, measured by homeless shelter contacts, was the exposure. The outcome was any diagnosis of a skin disorder and specific skin disorders recorded in the Danish National Patient Register. Information on diagnostic consultation type (i.e. dermatological, nondermatological and emergency room) and dermatological prescriptions was studied. We estimated adjusted incidence rate ratio (aIRR) (adjusted for sex, age and calendar year) and cumulative incidence.

Results
In total, 5 054 238 individuals (50.6% female) were included in the study population, accounting for 73 477 258 person-years at risk, with a start mean (SD) age of 39.4 (21.1) years. Of the total number of individuals, 759 991 (15.0%) received a skin diagnosis and 38 071 (0.7%) experienced homelessness. A 2.31-times [95% confidence interval (CI) 2.25–2.36] higher IRR of any diagnosed skin condition was associated with homelessness, higher for nondermatological and emergency room consultations. Homelessness was associated with a reduced IRR of a skin neoplasm diagnosis (aIRR 0.76, 95% CI 0.71–8.82) compared with no homelessness. By the end of follow-up, 2.8% (95% CI 2.5–3.0) of individuals experiencing homelessness had a skin neoplasm diagnosis vs. 5.1% (95% CI 4.9–5.3) of individuals not experiencing homelessness. Five or more shelter contacts during the first year from first contact was associated with the highest aIRR of any diagnosed skin condition (7.33, 95% CI 5.57–9.65) compared with no contacts.

Conclusions
Individuals experiencing homelessness have high rates of most diagnosed skin conditions, but a lower occurrence of skin cancer diagnosis. Diagnostic and medical patterns for skin disorders differed clearly between people experiencing homelessness and individuals without these experiences. The time after first homeless shelter contact is an important window of opportunity for mitigating and preventing skin disorders.