Abstract
Background
The experience of homelessness is a growing problem in the USA. Environmental exposures, chronic medical conditions and barriers to personal hygiene could increase risk of dermatological disease.
Objectives
To evaluate the association of homelessness with receiving a dermatological diagnosis and to compare dermatological diagnoses between unhoused and housed individuals.
Methods
We conducted a cross-sectional study of the ‘All of Us’ database, a national dataset from the USA with detailed participant-reported housing information. We constructed logistic regression models with dermatological diagnosis as binary outcome, housing status as binary exposure, and age, gender, race and ethnicity, geographical region, educational attainment, serious mental illness, alcohol-use disorder, and cocaine, opioid or stimulant drug use as confounders.
Results
Among 257 701 participants, people experiencing homelessness had lower odds of being diagnosed with any dermatological disease [adjusted odds ratio (aOR) 0.68, 95% confidence interval (CI) 0.63–0.72; P < 0.001]. Of participants with at least one dermatological diagnosis, people experiencing homelessness had greater odds of being diagnosed with bacterial skin infections (aOR 2.20, 95% CI 1.5–2.50; P < 0.001) and ectoparasitic disease (aOR 3.12, 95% CI 2.30–4.17; P < 0.001) but lower odds of all other diseases.
Conclusions
Our results underscore the need to improve access to care and water, sanitation and hygiene services for people experiencing homelessness to reduce risk of bacterial skin infection and ectoparasitic infestations.
Environmental exposures, chronic medical conditions and barriers to personal hygiene could increase risk of dermatologic disease in people experiencing homelessness. This article compares dermatologic diagnoses between unhoused and housed individuals. Results underscore the need to improve access to care and water, sanitation, and hygiene services for people experiencing homelessness to reduce risk of bacterial skin infection and ectoparasitic infestations.
What is already known about this topic?
Environmental exposures, chronic medical conditions and barriers to personal hygiene maintenance can increase the risk of dermatological disease in people experiencing homelessness.
What does this study add?
This study is the first national study from the USA to compare dermatological disease diagnoses in people experiencing homelessness to the general population.
Homelessness is a growing problem in the USA. On one night in 2024, 771 480 people were experiencing homelessness – the highest number reported since 2007 when the federal government began collecting these data. This represents a 19% increase in the number of people experiencing homelessness (PEH), including individuals and families with children, since 2007 and an 18% increase since 2023. In 2024, 35.5% of PEH were experiencing unsheltered homelessness in places not intended as a regular sleeping accommodation for humans (e.g. streets, vehicles and parks) and 64.4% were experiencing sheltered homelessness (e.g. staying in shelters and transitional housing).1 The detrimental health effects of homelessness are well documented,2–9 including its impact on skin health.7,10–13 In a nationwide Danish study, PEH had higher rates of any dermatological diagnosis compared with housed individuals.14 Multiple factors contribute to dermatological disease burden in PEH. These include exposure to ultraviolet radiation, toxic environmental exposures, damp clothing, improper footwear, long periods of standing, overcrowding and barriers to maintaining personal hygiene.15 Common chronic medical conditions such as hypertension and cardiovascular disease can worsen skin health, particularly on the lower extremities, where oedema further complicates skin injury from trauma, venous stasis or infections.10,13 High prevalence of substance-use disorders in PEH contributes to the increased risk of skin and soft tissue infections associated with injection drug use.10,16,17
Among studies evaluating skin conditions in PEH in the USA, almost all have been single-centre studies of PEH receiving dermatological outreach or consultation.18–25 None of these studies has examined how sheltered status, geographical location and educational attainment may impact dermatological disease diagnoses in PEH. Individuals experiencing unsheltered homelessness may have poor access to water, sanitation and hygiene (WASH) resources.26 Individuals experiencing sheltered homelessness may be staying in crowded living conditions. PEH in different geographical locations may have different environmental exposures or experience varied levels of social and healthcare support.27,28 Higher educational attainment may facilitate navigation of the healthcare system.29
We evaluated the association of experiencing homelessness with receipt of a dermatological diagnosis and compared dermatological diagnoses between PEH and housed individuals using the ‘All of Us’ database, a national dataset that uniquely includes detailed participant-reported housing information enabling categorization between unsheltered and sheltered homelessness. We hypothesized that PEH experience greater odds of dermatological disease diagnoses compared with housed individuals. Among PEH, we hypothesized that the odds of dermatological disease diagnosis would be higher in unsheltered PEH compared with sheltered PEH, and that geographical region and educational attainment would modify the effect of experiencing homelessness on receipt of a dermatological diagnosis.
Materials and methods
Data source
We used data from the National Institutes of Health’s All of Us Research Program’s Controlled Tier Dataset (v7), available to authorized users on the Researcher Workbench. The All of Us Research Program is a nationwide precision medicine initiative led by the National Institutes of Health that aims to build a diverse cohort of at least 1 million US participants to advance individualized prevention, treatment and care. The programme intentionally oversamples historically under-represented groups in biomedical research, including minority racial and ethnic groups, LGBTQ+ individuals and persons of lower socioeconomic status. Participants are enrolled through healthcare provider organizations, including federally qualified health centres, and community-based recruitment efforts, including mobile buses parked at strategic locations identified by community partners. Participants contribute survey responses, physical measurements and biospecimens.30,31 Participants may consent to share their electronic health record (EHR) data, which permits access to all prior and future healthcare visits at sites with a compatible EHR system.31 The All of Us Research Program was granted ethics approval from the All of Us Institutional Review Board.30 All data made available to researchers are provided without direct patient identifiers.32
Study design and population
We conducted a cross-sectional study to examine associations between housing status and dermatological diagnoses. We included participants from the All of Us database with EHR data, exposure and covariate data available (Figure S1; see Supporting Information).
Exposure
We characterized housing status from the baseline survey data completed by participants at time of enrollment. Participants first answered the question ‘Do you own or rent the place where you live?’ Answer options included ‘Own’, ‘Rent’, ‘Other Arrangement’ and ‘Prefer Not To Answer’. Those who answered ‘Other Arrangement’ were then asked the follow-up question ‘Where are you currently living?’ Answer choices for this question included ‘College Campus’, ‘Friend’, ‘Family’, ‘Motel/Hotel’, ‘Temporary Institute’, ‘Residential Facility’, ‘Transitional’, ‘Shelter’, ‘Outside’, ‘Other’ and ‘Skip’. We excluded participants who answered ‘Prefer Not To Answer’, ‘Skip’ and ‘Other’. Participants who answered ‘Motel/Hotel’, ‘Temporary Institute’, ‘Transitional’, ‘Shelter’ or ‘Outside’ were categorized as PEH. The remaining participants were categorized as being housed.
We further categorized PEH into those experiencing unsheltered vs. sheltered homelessness. Those who selected ‘outside’ were categorized as experiencing unsheltered homelessness; the remainder were categorized as experiencing sheltered homelessness.
Outcomes
Our primary outcomes were: (i) being diagnosed with any dermatological disease and (ii) being diagnosed with specific categories of dermatological disease, as defined below. We identified dermatological diagnoses by International Classification of Diseases, Ninth Revision (ICD-9) or Tenth Revision (ICD-10) codes, and the Systemized Nomenclature of Medicine Clinical Terms (SNOMED CT) codes. We grouped dermatological diagnostic codes into 28 categories based on group consensus (R.H., S.H., A.C.) (Table S1; see Supporting Information). For each disease category, we determined whether each participant had received one or more associated diagnoses within 1 year of when the housing data were reported (i.e. we searched 12 months prior and 12 months after). We chose this search range because PEH often transition in and out of homelessness, and most experience homelessness for less than 12 months at a time.1 If a participant did not have any diagnosis from a specific disease category within that timeframe, we considered them as not being affected by that disease category. If a participant had more than one diagnosis, all were included.
Covariates
Based on literature review, we included the following variables as confounders (Figure S2; see Supporting Information): age, gender, race and ethnicity, geographical region, educational attainment, serious mental illness, alcohol-use disorder, cocaine, opioid or stimulant drug use. We used calculated age based on time of housing survey completion and date of birth. We used self-reported data from participants at time of enrollment for gender (male, female, transgender), race (White, Asian, Black or African American, Middle Eastern or North African, more than one race, Native Hawaiian or Other Pacific Islander, and none of these), ethnicity (Hispanic or not Hispanic) and educational attainment (less than high school, high school graduate, more than high school). We combined race and ethnicity into one race and ethnicity variable by first grouping based on Hispanic vs. not Hispanic ethnicity, next defining non-Hispanic racial categories as per the source data, and then defining a ‘non-Hispanic Other’ racial category, as detailed in Table 1. We determined the presence of a diagnosis of serious mental illness (schizophrenia spectrum disorder, bipolar disorder, major depression, post-traumatic stress disorder), alcohol-use disorder, and cocaine, opioid or stimulant drug use disorder based on ICD-9, ICD-10 and SNOMED CT codes using the same approach we used in determining the outcome (Table S1).
Table 1.
Participant characteristics of study population stratified by housing status and the presence of a dermatological diagnosis
| All participants (n = 257 701) | Dermatological diagnosis only (n = 68 763) | Unhoused participants only (n = 6792) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Housed (n = 250 909) | Unhoused (n = 6792) | P-value | Housed (n = 67 455) | Unhoused (n = 1308) | P-value | No disease (n = 5484) | ≥ 1 disease (n = 1308) | P-value | |
| Age (years), mean (SD) | 52.5 (17.0) | 46 (12.2) | <0.001 | 55.8 (16.7) | 45.8 (12.3) | <0.001 | 46.0 (12.1) | 45.8 (12.3) | 0.60 |
| Age groups | <0.001 | <0.001 | <0.001 | ||||||
| 18–24 years | 12 350 (4.9) | 209 (3.1) | 2071 (3.1) | 30 (2.3) | 179 (3.3) | 30 (2.3) | |||
| 25–34 years | 35 108 (14.0) | 1134 (16.7) | 7323 (10.9) | 227 (17.4) | 907 (16.5) | 227 (17.4) | |||
| 35–44 years | 34 879 (13.9) | 1495 (22.0) | 8175 (12.1) | 327 (25.0) | 1168 (21.3) | 327 (25.0) | |||
| 45–54 years | 40 219 (16.0) | 1931 (28.4) | 10 233 (15.2) | 338 (25.8) | 1593 (29.1) | 338 (25.8) | |||
| 55–64 years | 54 291 (21.7) | 1703 (25.1) | 14 735 (21.8) | 308 (23.5) | 1395 (25.4) | 308 (23.6) | |||
| 65 years or older | 74 062 (29.5) | 320 (4.7) | 24 918 (36.9) | 78 (6.0) | 242 (4.4) | 78 (6.0) | |||
| Gender | <0.001 | <0.001 | 0.007 | ||||||
| Male | 93 077 (37.1) | 4581 (67.4) | 24 406 (36.2) | 836 (63.9) | 3745 (68.3) | 836 (63.9) | |||
| Female | 156 809 (62.5) | 2158 (31.8) | 42 781 (63.4) | ≤465 (≤36) | ≤1700 (≤31) | ≤465 (≤36) | |||
| Transgender male or female | 1023 (0.4) | 53 (0.8) | 268 (0.4) | ≤20 (≤2) | ≤50 (≤1) | ≤20 (≤2) | |||
| Race and ethnicitya | <0.001 | <0.001 | <0.001 | ||||||
| White | 140 101 (55.8) | 2695 (39.7) | 43 897 (65.1) | 678 (51.8) | 2017 (36.8) | 678 (51.8) | |||
| Asian | 7550 (3.0) | 29 (0.4) | 1606 (2.4) | ≤20 (≤0.5) | ≤30 (≤0.6) | ≤20 (≤0.5) | |||
| Black or African American | 47 184 (18.8) | 2711 (39.9) | 9567 (14.2) | 344 (26.3) | 2367 (43.2) | 344 (26.3) | |||
| Hispanic | 47 909 (19.1) | 1075 (15.8) | 10 298 (15.3) | 228 (17.4) | 847 (15.4) | 228 (17.4) | |||
| Otherb | 8165 (3.3) | 282 (4.2) | 2087 (3.1) | ≤55 (≤4.5) | ≤230 (≤4.5) | ≤55 (≤4.5) | |||
| Geographical regionc | <0.001 | <0.001 | <0.001 | ||||||
| Midwest | 59 808 (23.8) | 1167 (17.2) | 17 060 (25.3) | 155 (11.9) | 1012 (18.5) | 155 (11.9) | |||
| Northeast | 77 672 (31.0) | 1938 (28.5) | 24 581 (36.4) | 451 (34.5) | 1487 (27.1) | 451 (34.5) | |||
| South | 42 973 (17.1) | 2040 (30.0) | 10 184 (15.1) | 219 (16.7) | 1821 (33.2) | 219 (16.7) | |||
| West | 70 456 (28.1) | 1647 (24.2) | 15 630 (23.2) | 483 (36.9) | 1164 (21.2) | 483 (36.9) | |||
| Education | <0.001 | <0.001 | 0.15 | ||||||
| Less than high school | 23 174 (9.2) | 1635 (24.1) | 4809 (7.1) | 309 (23.6) | 1326 (24.2) | 309 (23.6) | |||
| High school graduate | 47 823 (19.1) | 2837 (41.8) | 11 098 (16.5) | 523 (40.0) | 2314 (42.2) | 523 (40.0) | |||
| More than high school | 179 912 (71.7) | 2320 (34.2) | 51 548 (76.4) | 476 (36.4) | 1884 (34.4) | 476 (36.4) | |||
| Cocaine, opioid or stimulant-use disorder | 8145 (3.2) | 1271 (18.7) | <0.001 | 3683 (5.5) | 531 (40.6) | <0.001 | 740 (13.5) | 531 (40.6) | <0.001 |
| Alcohol-use disorder | 6635 (2.6) | 792 (11.7) | <0.001 | 2694 (4.0) | 278 (21.3) | <0.001 | 514 (9.4) | 278 (21.3) | <0.001 |
| Serious mental illness | 42 123 (16.8) | 1647 (24.2) | <0.001 | 19 270 (28.6) | 634 (48.5) | <0.001 | 1013 (18.5) | 634 (48.5) | <0.001 |
Data are presented as n (%) unless otherwise specified. χ2 tests and t-tests were used to test for significant differences, as appropriate. The Benjamini–Hochberg method was used separately for the three groups (all participants, dermatological diagnosis only and unhoused participants) to account for multiple testing. Cell counts <20 are not reported in adherence with the All of Us data use policy. Other cells are suppressed to prevent back-calculation. aAll categories other than ‘Hispanic’ indicated non-Hispanic ethnicity. bOther includes Middle Eastern or North African, ‘More than one population’, Native Hawaiian or Other Pacific Islander, and ‘None of these’. cUS regions as categorized by Census Bureau. Midwest includes Indiana, Illinois, Michigan, Ohio, Wisconsin, Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, South Dakota. Northeast includes Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont, New Jersey, New York, Pennsylvania. South includes Delaware, DC, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, West Virginia, Alabama, Kentucky, Mississippi, Tennessee, Arkansas, Louisiana, Oklahoma, Texas. West includes Arizona, Colorado, Idaho, New Mexico, Montana, Utah, Nevada, Wyoming, Alaska, California, Hawaii, Oregon, Washington.
Statistical analysis
We used χ2, Fisher’s exact and t-tests to evaluate dependence between sociodemographic characteristics, housing status and the dermatological diagnosis outcome. For our primary analyses, we first fitted a logistic regression model for dermatological diagnosis in the entire study population using PEH vs. housed as the primary exposure, and age, gender, race and ethnicity, geographical region, educational attainment, serious mental illness, alcohol-use disorder, cocaine, opioid or stimulant drug use as confounders. We fitted similar logistic regression models for separate dermatological disease categories, using the subpopulation restricted to include individuals with at least one diagnosis. For each model, the outcome was the binary indicator of presence of that dermatological disease. Adjustment variables included those mentioned above. Models were estimated using complete cases as the level of missingness in included variables was deemed acceptable (all covariates <5%; housing status: 6.2%) (Figure S1).
For secondary analyses focusing on sheltered status, we first fitted a logistic model in the subpopulation experiencing homelessness using sheltered homelessness vs. unsheltered homelessness as the exposure, receipt of a dermatological diagnosis as the outcome, and adjusting for confounders mentioned previously. Then, we repeated this analysis for selected dermatological outcomes in the subpopulation experiencing homelessness and with at least one dermatological diagnosis. For each resulting model, the exposure was sheltered homelessness vs. unsheltered homelessness, the outcome was one of five dermatological disease categories (bacterial skin infection, cutaneous malignancy, ectoparasitic disease, ulcer or wound, ultraviolet radiation-related skin condition) and adjusted for confounders mentioned previously. We selected these five disease categories based on the hypothesis that unsheltered PEH may have increased exposure to environmental factors, and this may increase the likelihood of the five selected disease categories in unsheltered PEH, as compared with sheltered PEH. As an additional secondary analysis, we examined whether geographical region and educational attainment were effect modifiers of housing status (PEH/housed) for bacterial infection and ectoparasitic disease. Models adjusted for confounders mentioned previously. Likelihood-ratio tests were used to evaluate significance of hypothesized effect modifiers represented as two-way interactions in regression models.
All analyses were conducted in the R environment provided by the All of Us Research Workbench.30 We used the Benjamini–Hochberg method to adjust P-values for multiple hypothesis testing. P-values <0.05 were considered significant. Data were accessed and analysed between 17 April 2023 and 6 September 2024.
Results
Group characteristics
A total of 257 701 participants were included in this study (Figure S1), of whom 2.6% (n = 6792) were PEH (Table 1). Compared with the housed population, PEH were younger [mean (SD) age 46 (12.2) vs. 52.5 (17) years; P < 0.001], a greater proportion identified as male (67.4% vs. 37.1%; P < 0.001), a greater proportion identified as Black/African American (39.9% vs. 18.8%; P < 0.001) and a smaller proportion had more than a high school education (34.2% vs. 71.7%; P < 0.001). The largest proportion of PEH were from the South, followed by Northeast, West and Midwest. Serious mental illness and substance-use disorders were more frequently diagnosed in PEH.
In total, 68 763 participants had at least one dermatological diagnosis within the study period, representing 26.7% of the total study population. Of these 68 763 participants, 1.9% n = 1308 were PEH. Compared with the 67 455 housed participants with a dermatological diagnosis, PEH with a dermatological diagnosis were younger [mean (SD) age 45.8 (12.3) vs. 55.8 (16.7) years; P<0.001] and predominantly male (63.9% vs. 36.2%; P<0.001). A greater proportion of PEH identified as Black/African American (26.3% vs 14.2%; P<0.001) or Hispanic (17.4% vs. 15.3%; P<0.001), and a smaller proportion had more than a high school education (36.4% vs. 76.4%; P<0.001). The largest proportion of PEH with a dermatological diagnosis was from the West, followed by the Northeast, South, and Midwest.
Odds of dermatological disease diagnosis
In the total study population, PEH had lower odds of being diagnosed with any dermatological disease than housed individuals [adjusted odds ratio (aOR) 0.68, 95% confidence interval (CI) 0.63–0.72; P < 0.001] (Table 2). In the subset of the population with at least one dermatological diagnosis, PEH had greater odds of being diagnosed with bacterial skin infections (aOR 2.20, 95% CI 1.5–2.50; P < 0.001) and ectoparasitic disease (aOR 3.12, 95% CI 2.30–4.17; P < 0.001) but had lower odds of other dermatological conditions, such as cutaneous malignancy (aOR 0.26, 95% CI 0.17–0.38; P < 0.001) (Table 2).
Table 2.
Unadjusted and adjusted odds of dermatological diagnosis stratified by housing status
| Housed (n = 250 909) | Unhoused (n = 6792) | OR, unadjusted (95% CI) | P-value | OR, adjusted (95% CI) | P-value | |
|---|---|---|---|---|---|---|
| Any dermatological disease | 67 455 (26.9) | 1308 (19.3) | 0.65 (0.61–0.69) | <0.001 | 0.68 (0.63–0.72) | <0.001 |
| Category | Housed (n = 67 455) | Unhoused (n = 1308) | OR, unadjusted (95% CI) | P-value | OR, adjusted (95% CI) |
P-value |
|---|---|---|---|---|---|---|
| Acneiform/follicular disorder | 9941 (14.7) | 110 (8.4) | 0.53 (0.43–0.64) | <0.001 | 0.64 (0.52–0.79) | <0.001 |
| Alopecia | 3498 (5.2) | ≤20 (≤1.1) | 0.20 (0.11–0.32) | <0.001 | 0.33 (0.19–0.55 | <0.001 |
| Bacterial skin infection | 15 030 (22.3) | 768 (58.7) | 4.96 (4.44–5.55) | <0.001 | 2.20 (1.95–2.50) | <0.001 |
| Benign growth or mass | 23 058 (34.2) | 79 (6.0) | 0.12 (0.10–0.15) | <0.001 | 0.32 (0.25–0.40) | <0.001 |
| Bullous/vesicular/acantholytic disease | 329 (0.5) | ≤20 (≤0.1) | 0.16 (0.01–0.69) | 0.10 | 0.28 (0.02–1.26) | 0.27 |
| Cutaneous malignancy | 12 896 (19.1) | 25 (1.9) | 0.08 (0.05–0.12) | <0.001 | 0.26 (0.17–0.38) | <0.001 |
| Disorders of keratinization | 3246 (4.8) | 80 (6.1) | 1.29 (1.02–1.61) | 0.05 | 1.09 (0.85–1.38) | 0.55 |
| Ectoparasitic disease | 479 (0.7) | 63 (4.8) | 7.1 (5.36–9.18) | <0.001 | 3.12 (2.30–4.17) | <0.001 |
| Erythema | 1502 (2.2) | 26 (2.0) | 0.89 (0.59–1.29) | 0.61 | 0.73 (0.48–1.07) | 0.18 |
| Disorder of sweat glands | 1198 (1.8) | ≤20 (≤0.3) | 0.17 (0.05–0.40) | <0.001 | 0.28 (0.09–0.65) | 0.02 |
| Drug reaction | 689 (1.0) | ≤20 (≤0.9) | 0.90 (0.48–1.52) | 0.74 | 0.73 (0.38–1.26) | 0.35 |
| Dyspigmentation | 9945 (14.7) | 35 (2.7) | 0.16 (0.11–0.22) | <0.001 | 0.38 (0.26–0.52) | <0.001 |
| Eczematous dermatitis | 18 466 (27.4) | 202 (15.4) | 0.48 (0.42–0.56) | <0.001 | 0.59 (0.50–0.68) | <0.001 |
| Fungal/mycobacterial skin infection | 3039 (4.5) | 45 (3.4) | 0.76 (0.55–1.01) | 0.10 | 0.79 (0.57–1.06) | 0.19 |
| Nail disorder | 6084 (9.0) | 102 (7.8) | 0.85 (0.69–1.04) | 0.18 | 0.90 (0.72–1.10) | 0.35 |
| Other | 3217 (4.8) | ≤20 (≤0.3) | 0.28 (0.17–0.43) | <0.001 | 0.44 (0.26–0.68) | 0.001 |
| Papulosquamous disorder | 5078 (7.5) | 40 (3.1) | 0.39 (0.28–0.52) | <0.001 | 0.55 (0.39–0.75) | <0.001 |
| Pruritus/excoriation disorder | 6288 (9.3) | 120 (9.2) | 0.98 (0.81–1.18) | 0.87 | 0.90 (0.73–1.09) | 0.35 |
| Rheumatological disease | 917 (1.4) | ≤20 (≤0.7) | 0.50 (0.24–0.91) | 0.07 | 0.66 (0.32–1.22) | 0.30 |
| Ulcer/wound | 4060 (6.0) | 115 (8.8) | 1.51 (1.23–1.82) | <0.001 | 0.70 (0.56–0.85) | 0.001 |
| Urticarial disease | 2468 (3.7) | 35 (2.7) | 0.72 (0.51–1.00) | 0.10 | 0.81 (0.56–1.12) | 0.29 |
| Ultraviolet radiation-related skin condition | 4780 (7.1) | ≤20 (≤1.4) | 0.18 (0.11–0.28) | <0.001 | 0.42 (0.25–0.65) | <0.001 |
| Vasculitis/vasculopathy | 124 (0.2) | ≤20 (≤0.3) | 1.25 (0.31–3.30) | 0.74 | 1.03 (0.25–2.92) | 0.96 |
| Viral skin infection | 1133 (1.7) | ≤20 (≤1.2) | 0.72 (0.42–1.15) | 0.27 | 0.84 (0.49–1.36) | 0.57 |
Data are presented as n (%) unless otherwise specified. Logistic regression models were developed to estimate odds ratios (ORs) and adjusted ORs. The following variables were included as possible confounders: age at the time of housing survey completion, gender (male, female, transgender), race and ethnicity (Hispanic, Non-Hispanic White, Non-Hispanic Asian, Non-Hispanic Black or African American, Other), geographical region (West, Midwest, South, Northeast), educational attainment (less than high school, high school graduate, more than high school), serious mental illness, alcohol-use disorder, and cocaine, opioid or stimulant-use disorder. The Benjamini–Hochberg method was used to adjust P-values for multiple testing of associations by dermatological disease diagnosis. Neutrophilic dermatoses, granulomatous disease, dermal atrophy and nutrient deficiency were excluded because one or more of the cell sizes were 0. Cell counts <20 are not reported in adherence with the All of Us data use policy. CI, confidence interval.
Among PEH, people experiencing unsheltered homelessness (n = 1078) had greater odds of having any dermatological diagnosis (aOR 1.2, 95% CI 1.01–1.42; P = 0.04). Among PEH who had at least one dermatological diagnosis, people experiencing unsheltered homelessness (n = 267) had higher odds of being diagnosed with bacterial skin infections (aOR 1.42, 95% CI 1.02–1.99; P = 0.04) (Table 3).
Table 3.
Unadjusted and adjusted odds ratios in persons experiencing homelessness comparing select dermatological disease outcome odds by sheltered status
| OR, unadjusted (95% CI) | P-value | OR, adjusted (95% CI) | P-value | |
|---|---|---|---|---|
| Any dermatological disease | 1.48 (1.26–1.72) | <0.001 | 1.20 (1.01–1.42) | 0.04 |
| Category | ||||
| Bacterial skin infection | 2.37 (1.76–3.22) | <0.001 | 1.42 (1.02–1.99) | 0.04 |
| Cutaneous malignancy | 0.74 (0.21–1.96) | 0.58 | 0.84 (0.23–2.43) | 0.77 |
| Ectoparasitic disease | 0.40 (0.15–0.86) | 0.03 | 0.56 (0.21–1.26) | 0.20 |
| Ulcer/wound | 1.22 (0.76–1.89) | 0.39 | 0.88 (0.54–1.42) | 0.62 |
| Ultraviolet radiation-related skin condition | 0.77 (0.18–2.38) | 0.68 | 0.55 (0.12–1.81) | 0.38 |
Sheltered status was determined by participant-reported information in the housing survey. Logistic regression models were developed to estimate odds ratios (ORs) and adjusted ORs. The latter adjusted for the variables listed in Table 2. The Benjamini–Hochberg method was used to account for multiple testing. CI, confidence interval.
In our effect modification analysis of PEH diagnosed with bacterial skin infection, adjusted diagnosis ORs comparing unhoused with housed individuals varied significantly across geographical regions and educational attainment (P < 0.001), with the highest estimate observed in the West (aOR 3.14, 95% CI 2.51–3.94) and among individuals with more than a high school education (aOR 3.00, 95% CI 2.47–3.66) (Table 4). Similarly, among PEH diagnosed with an ectoparasitic infestation, adjusted diagnosis ORs comparing unhoused with housed individuals varied significantly across geographical regions and educational attainment (P < 0.001), with the lowest estimate observed in the West (aOR 1.04, 95% CI 0.50–1.92) and among individuals with less than a high school education (aOR 1.78, 95% 0.93–3.16) (Table 4).
Table 4.
Effect modification by geographical region and educational attainment on odds of dermatological diagnosis
| Categories | OR, adjusted (95% CI) | |
|---|---|---|
| Bacterial skin infection | Unhoused | |
| Geographical region | ||
| Northeast | 1.98 (1.62–2.41) | |
| Midwest | 1.56 (1.11–2.18) | |
| South | 1.92 (1.45–2.56) | |
| West | 3.14 (2.51–3.94) | |
| Educational attainment | ||
| More than high school | 3.00 (2.47–3.66) | |
| High school graduate | 1.73 (1.43–2.09) | |
| Less than high school | 2.00 (1.56–2.58) | |
| Ectoparasitic disease | Geographical region | |
| Northeast | 5.54 (3.55–8.40) | |
| Midwest | 5.35 (2.53–10.19) | |
| South | 3.33 (1.76–5.88) | |
| West | 1.04 (0.50–1.92) | |
| Educational attainment | ||
| More than high school | 6.81 (4.34–10.29) | |
| High school graduate | 2.33 (1.44–3.64) | |
| Less than high school | 1.78 (0.93–3.16) |
Estimated odds ratios (ORs) evaluating the presence of a multiplicative interaction between housing status and either geographical region or education attainment level while controlling for the following confounders: age at the time of housing survey completion, gender (male, female, transgender), race and ethnicity (Hispanic, Non-Hispanic White, Non-Hispanic Asian, Non-Hispanic Black or African American, Other), geographical region (West, Midwest, South, Northeast), educational attainment (less than high school, high school graduate, more than high school), serious mental illness, alcohol-use disorder, and cocaine, opioid or stimulant-use disorder. Interaction significance for each two-way comparison was assessed using the likelihood ratio test for inclusion of the interaction terms (significance level P < 0.05). Estimates are provided for two diagnoses that yielded significant results. CI, confidence interval.
Discussion
In this study, we found that PEH had lower odds of a dermatological diagnosis. Among participants with at least one dermatological diagnosis, PEH were more likely to be diagnosed with bacterial skin infections and ectoparasitic disease. Among PEH with a dermatological diagnosis, unsheltered PEH had higher odds of a bacterial skin infection diagnosis compared with sheltered PEH. We also observed effect modification by educational attainment on the association between housing status and receipt of a diagnosis of bacterial skin infection and ectoparasitic disease, as well as effect modification by geographical region on the association between housing status and receipt of a bacterial skin infection diagnosis.
Our finding of lower odds of a dermatological diagnosis in PEH contrasts with a Danish nationwide, population-based longitudinal cohort study that found homelessness was associated with a 2.31 times (95% CI 2.25–2.36) higher incidence rate ratio of any diagnosed skin condition.14 This discrepancy in study findings is most likely due to differences in the study population and differences between the healthcare system in Denmark and the USA. The Danish study utilized comprehensive, population-level data from national health and administrative registries within a public healthcare system that offers essential health services for free at point-of-care, ensuring reliable capture of diagnoses across healthcare settings and with reduced concern that individuals in need of healthcare were unable to access it. In contrast, the All of Us dataset – while large and diverse – is not a population-based sample and is more susceptible to underascertainment of diagnoses among populations with limited healthcare access, such as PEH. Differential access to care in the US healthcare system may result in fewer dermatological diagnoses being made in PEH, rather than a decreased risk of dermatological disease. There is a large body of work from across the USA describing the barriers that PEH experience in obtaining healthcare: limited clinic hours, distance to clinic, health insurance and low health literacy, and less-measurable barriers such as bias of clinicians.33–43 Another explanation is that PEH are presenting to care but dermatological conditions are underdiagnosed, potentially due to not being a priority concern for the patient or the healthcare team. Differences in healthcare utilization patterns and coding practices between the USA and Denmark may also contribute to the contrasting findings.
One additional explanation is that the most medically vulnerable among the unhoused population may not be represented in our dataset. Unsheltered PEH are underrepresented in our study. Of the unhoused participants in our study, 15.9% were unsheltered vs. 35.5% of the total US population experiencing homelessness.1 Non-traditional healthcare delivery models, such as direct outreach (e.g. street medicine) or the integration of healthcare services within navigation centres, help to facilitate access to care for PEH.19,33 The All of Us Research Program may not be recruiting from or near nontraditional healthcare delivery sites where PEH are accessing healthcare services or these sites may not be contributing electronic health record data. Additionally, even with the use of outreach programmes, unsheltered PEH may still be unconnected to healthcare.44
Among participants with a dermatological diagnosis, PEH were more likely to have bacterial skin infection and ectoparasitic disease, which is consistent with existing studies.7,13,14,21,23,45–47 When examining the influence of geographical region, we found that PEH living in the West were more likely to receive a diagnosis of bacterial skin infection than PEH living in other regions. This finding may partially reflect the higher proportion of unsheltered homelessness in western states, such as California, where over two-thirds of PEH are unsheltered.1 Possible explanations for the increased likelihood of bacterial skin infection among unsheltered PEH include environmental exposures leading to skin trauma resulting in inoculation of bacteria coupled with difficulty accessing WASH services.26 Our finding that unsheltered PEH had higher odds of bacterial skin infection than sheltered PEH supports this interpretation. However, we acknowledge that unsheltered homelessness is also common in the Southern USA, yet the odds of bacterial skin infection diagnoses were not as elevated in that region in our dataset.1 Regional variation in healthcare access for PEH may also contribute to the observed regional disparity. The West has more federally qualified health centres per capita (4.84 sites per 100 000), than the South (3.49 sites per 100 000).48 These federally qualified health centres are an essential component of the safety net health system that facilitate better access to care among PEH, and in the context of this study, may increase opportunities for diagnoses to be captured in this dataset.
Our finding of lower odds of cutaneous malignancy diagnosis aligns with a Danish cohort study that found a lower rate of skin cancer diagnosis in PEH.14 In contrast, one study at a free clinic in the USA found increased rates of skin cancer diagnosis in their population of PEH vs. their housed counterparts.18 Patients in this study had all received full-body skin exams. As skin cancers may only be identified incidentally on full-body skin exams by a dermatologist or primary care provider,49–51 ascertainment bias may have influenced the results of our study and the Danish cohort study, as it is unlikely that all study participants received full-body skin exams.
We found that PEH with higher educational attainment were more likely to receive a diagnosis of bacterial skin infection or ectoparasite infestation. Higher educational attainment may reflect better health literacy and the ability to navigate healthcare systems to receive a diagnosis and treatment.
Our study has limitations. Study findings may have limited generalizability as the study sample is not nationally representative or population based. Selection bias is likely present as the study sample includes individuals who enrolled in the All of Us Research Program, accessed healthcare at a participating site and provided consent for their EHR information to be accessed for research. Unsheltered PEH are under-represented in the study sample, as discussed earlier. Misclassification bias may also be present that differentially impacts PEH. We do not know which specialty made the dermatological diagnoses, and we were unable to account for medical encounter type such as inpatient, outpatient or emergency department. We are unable to assess disease severity. Our definition of homelessness is based on housing status at one point in time. We could not determine the duration of homelessness, number of episodes of homelessness or changes in sheltered status, which may impact odds of having skin disease and type of skin disease. We cannot draw causal inference from this study due to its cross-sectional design.
This is the first national study in the USA to examine dermatological disease among PEH. This study took advantage of a national dataset with detailed participant-reported housing information to advance our understanding of how sheltered status, educational attainment and geographical region may influence dermatological diagnoses among PEH, supports the possibility of underdiagnosis of skin malignancy among PEH, and underscores the need for public health interventions to improve skin health among PEH. Housing for all individuals in society should be a national, state and local priority. As work is done to achieve this, special attention should be paid to improve PEH access to care and WASH services to reduce the risk of bacterial skin infection and ectoparasitic infestations.
Supplementary Material
Acknowledgements
R.H., S.H. and D.C. had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. We gratefully acknowledge All of Us participants for their contributions, without whom this research would not have been possible. We also thank the National Institutes of Health’s All of Us Research Program for making available the participant data analysed in this study.
Contributor Information
Reneé Haughton, Department of Dermatology, University of California, Davis, Sacramento, CA, USA.
Derek Chen, Icahn School of Medicine at Mount Sinai, New York, NY, USA; Department of Family Medicine, Oregon Health and Science University, Portland, OR, USA.
Samantha Herbert, Department of Dermatology, University of California, Davis, Sacramento, CA, USA; University of Miami Miller School of Medicine, Miami, FL, USA.
Stephen Shiboski, Department of Epidemiology & Biostatistics, University of California, San Francisco, San Francisco, CA, USA.
Jinoos Yazdany, Department of Medicine, Division of Rheumatology, University of California, San Francisco School of Medicine, San Francisco, CA, USA.
Aileen Y Chang, Department of Dermatology, University of California, San Francisco School of Medicine, San Francisco, CA, USA; Department of Dermatology, Zuckerberg San Francisco General Hospital and Trauma Center, San Francisco, CA, USA.
Author contributions
Reneé Haughton (Conceptualization [equal], Data curation [equal], Investigation [equal], Project administration [equal], Resources [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Derek Chen (Conceptualization [equal], Data curation [equal], Formal analysis [equal], Investigation [equal], Methodology [equal], Project administration [equal], Resources [equal], Validation [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Samantha Herbert (Conceptualization [equal], Data curation [equal], Investigation [equal], Project administration [equal], Resources [equal], Writing—review & editing [equal]), Steve Shiboski (Methodology [equal], Supervision [equal], Writing—review & editing [equal]), Jinoos Yazdany (Supervision [equal], Writing—review & editing [equal]) and Aileen Y. Chang (Conceptualization [equal], Data curation [equal], Funding acquisition [lead], Investigation [equal], Methodology [lead], Project administration [lead], Resources [equal], Supervision [lead], Validation [lead], Visualization [lead], Writing—original draft [lead], Writing—review & editing [lead])
Conflicts of interest
The authors declare no conflicts of interest.
Funding sources
A.Y.C. reports support from the Dermatology Foundation Public Health Career Development Award and the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under award No. K23AR082918 during the conduct of this study. The National Institute of Arthritis and Musculoskeletal and Skin Diseases and the Dermatology Foundation had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review or approval of the manuscript; and decision to submit the manuscript for publication.
Data availability
This study is a secondary data analysis using controlled tier data from the National Institutes of Health’s All of Us Research Program. Authorized users may access this data.
Ethics statement
The All of Us Research Program was granted ethics approval from the All of Us Institutional Review Board.30 All data made available to researchers is provided without direct patient identifiers.32 This study was not considered human subjects research by the Institutional Review Boards at the University of California, Davis and Oregon Health & Science University.
Patient consent
Not applicable.
Supporting Information
Additional Supporting Information may be found in the online version of this article at the publisher’s website.
References
- 1. US Department of Housing and Urban Development . The 2024 Annual Homeless Assessment Report (AHAR) to Congress – Part 1: Point-in-Time Estimates of Homelessness. 2024. Available at: https://www.huduser.gov/portal/sites/default/files/pdf/2024-AHAR-Part-1.pdf (last accessed 30 July 2025).
- 2. Aldridge RW, Story A, Hwang SW et al. Morbidity and mortality in homeless individuals, prisoners, sex workers, and individuals with substance use disorders in high-income countries: a systematic review and meta-analysis. Lancet 2018; 391:241–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Huang C, Foster H, Paudyal V et al. A systematic review of the nutritional status of adults experiencing homelessness. Public Health 2022; 208:59–67. [DOI] [PubMed] [Google Scholar]
- 4. Figueiredo RL, Hwang SW, Quinonez C. Dental health of homeless adults in Toronto, Canada. J Public Health Dent 2013; 73:74–8. [DOI] [PubMed] [Google Scholar]
- 5. Lewer D, Aldridge RW, Menezes D et al. Health-related quality of life and prevalence of six chronic diseases in homeless and housed people: a cross-sectional study in London and Birmingham, England. BMJ Open 2019; 9:e025192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Baggett TP, Liauw SS, Hwang SW. Cardiovascular disease and homelessness. J Am Coll Cardiol 2018; 71:2585–97. [DOI] [PubMed] [Google Scholar]
- 7. Raoult D, Foucault C, Brouqui P. Infections in the homeless. Lancet Infect Dis 2001; 1:77–84. [DOI] [PubMed] [Google Scholar]
- 8. Fazel S, Geddes JR, Kushel M. The health of homeless people in high-income countries: descriptive epidemiology, health consequences, and clinical and policy recommendations. Lancet 2014; 384:1529–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gutwinski S, Schreiter S, Deutscher K et al. The prevalence of mental disorders among homeless people in high-income countries: an updated systematic review and meta-regression analysis. PLOS Med 2021; 18:e1003750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Moy JA, Sanchez MR. The cutaneous manifestations of violence and poverty. Arch Dermatol 1992; 128:829–39. [PubMed] [Google Scholar]
- 11. Coates SJ, Amerson EH, Chang AY. Dermatologic care of persons experiencing homelessness: key concepts in an era of housing instability. JAMA Dermatol 2020; 156:945–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Evans YN, Handschin SM, Giesel AE. Health care utilization in homeless youth. J Community Health 2014; 39:521–3. [DOI] [PubMed] [Google Scholar]
- 13. Adly M, Woo TE, Traboulsi D et al. Understanding dermatologic concerns among persons experiencing homelessness: a scoping review and discussion for improved delivery of care. J Cutan Med Surg 2021; 25:616–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Nilsson SF, Ali Z, Laursen TM et al. Association of homelessness and skin conditions: a Danish population-based cohort study. Br J Dermatol 2023; 188:760–9. [DOI] [PubMed] [Google Scholar]
- 15. Moffa M, Cronk R, Fejfar D et al. A systematic scoping review of environmental health conditions and hygiene behaviors in homeless shelters. Int J Hyg Environ Health 2019; 222:335–46. [DOI] [PubMed] [Google Scholar]
- 16. Bharel M, Lin WC, Zhang J et al. Health care utilization patterns of homeless individuals in Boston: preparing for Medicaid expansion under the Affordable Care Act. Am J Public Health 2013; 103:S311–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Lloyd-Smith E, Wood E, Zhang R et al. Risk factors for developing a cutaneous injection-related infection among injection drug users: a cohort study. BMC Public Health 2008; 8:405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Grossberg AL, Carranza D, Lamp K et al. Dermatologic care in the homeless and underserved populations: observations from the Venice Family Clinic. Cutis 2012; 89:25–32. [PubMed] [Google Scholar]
- 19. Rasul T, Mathew M, Morrison BW. Approaching the dermatologic care of people experiencing unsheltered homelessness through a volunteer street dermatology model. JAAD Int 2023; 11:52–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Truong A, Secrest AM, Zhang M et al. A survey of dermatologic health-related quality of life and resource access in patients experiencing homelessness. J Am Acad Dermatol 2021; 85:775–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Zakaria A, Amerson EH, Kim-Lim P et al. Characterization of dermatological diagnoses among hospitalized patients experiencing homelessness. Clin Exp Dermatol 2022; 47:117–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Contag C, Lowenstein S, Jain S et al. Survey of symptomatic dermatologic disease in homeless patients at a shelter-based clinic. Our Dermatol Online 2017; 8:133–7. [Google Scholar]
- 23. Zakaria A, Abuabara K, Szumowski J et al. Association between homelessness and group A streptococcus skin and soft tissue infections among hospitalized dermatology consult patients. JAMA Dermatol 2022; 158:321–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Stratigos AJ, Stern R, Gonzalez E et al. Prevalence of skin disease in a cohort of shelter-based homeless men. J Am Acad Dermatol 1999; 41:197–202. [DOI] [PubMed] [Google Scholar]
- 25. Truong A, Secrest AM, Fino NF et al. Homeless status and associations with treatment disparities for common skin conditions. Br J Dermatol 2020; 183:582–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Avelar Portillo LJ, Kayser GL, Ko C et al. Water, Sanitation, and Hygiene (WaSH) insecurity in unhoused communities of Los Angeles, California. Int J Equity Health 2023; 22:108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Tsai J, Jenkins D, Lawton E. Civil legal services and medical-legal partnerships needed by the homeless population: a national survey. Am J Public Health 2017; 107:398–401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Christian NJ, Havlik J, Tsai J. The use of mobile medical units for populations experiencing homelessness in the United States: a scoping review. J Gen Intern Med 2024; 39:1474–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Zimmerman E, Woolf SH. Understanding the relationship between education and health. Discussion Paper, Institute of Medicine, Washington, DC. 2014. Available at: https://nam.edu/wp-content/uploads/2015/06/BPH-UnderstandingTheRelationship1.pdf (last accessed 8 January 2026). [Google Scholar]
- 30. All Us Research Program Investigators . The “All of Us” research program. N Engl J Med 2019; 381:668–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. National Institutes of Health . All of Us Research Program: Operational Protocol. 2021. Available at: https://allofusassets.ppscaouprod.org/prod/pdf/All_of_Us_Research_Program_Operational_Protocol_2022.pdf (last accessed 29 August 2025).
- 32. All of Us Research Program . Framework for Access to All of Us Data Resources. 2021. Available at https://support.researchallofus.org/hc/enus/article_attachments/22307193818644 (last accessed 29 August 2025).
- 33. Zlotnick C, Zerger S, Wolfe PB. Health care for the homeless: what we have learned in the past 30 years and what’s next. Am J Public Health 2013; 103:S199–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Varley AL, Montgomery AE, Steward J et al. Exploring quality of primary care for patients who experience homelessness and the clinicians who serve them: what are their aspirations? Qual Health Res 2020; 30:865–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Whitley R. Fear and loathing in New England: examining the health-care perspectives of homeless people in rural areas. Anthropol Med 2013; 20:232–43. [DOI] [PubMed] [Google Scholar]
- 36. Henderson MD, McCurry IJ, Deatrick JA et al. Experiences of adult men who are homeless accessing care: a qualitative study. J Transcult Nurs 2022; 33:199–207. [DOI] [PubMed] [Google Scholar]
- 37. Franco A, Meldrum J, Ngaruiya C. Identifying homeless population needs in the emergency department using community-based participatory research. BMC Health Serv Res 2021; 21:428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Omerov P, Craftman AG, Mattsson E et al. Homeless persons’ experiences of health- and social care: a systematic integrative review. Health Soc Care Community 2020; 28:1–11. [DOI] [PubMed] [Google Scholar]
- 39. Meehan AA, Cox SN, Thuo NB et al. Previous health care experiences’ influence on health care perceptions among residents of six homeless shelters in Seattle, Washington, July–October 2021. J Patient Cent Res Rev 2023; 10:111–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Miller JP, Hutton J, Doherty C et al. A scoping review examining patient experience and what matters to people experiencing homelessness when seeking healthcare. BMC Health Serv Res 2024; 24:492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Thorndike AL, Yetman HE, Thorndike AN et al. Unmet health needs and barriers to health care among people experiencing homelessness in San Francisco’s Mission District: a qualitative study. BMC Public Health 2022; 22:1071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Hudson AL, Nyamathi A, Greengold B et al. Health-seeking challenges among homeless youth. Nurs Res 2010; 59:212–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Moore-Nadler M, Clanton C, Roussel L. Storytelling to capture the health care perspective of people who are homeless. Qual Health Res 2020; 30:182–95. [DOI] [PubMed] [Google Scholar]
- 44. Kopanitsa V, McWilliams S, Leung R et al. A systematic scoping review of primary health care service outreach for homeless populations. Fam Pract 2023; 40:138–51. [DOI] [PubMed] [Google Scholar]
- 45. Brouqui P, Raoult D. Arthropod-borne diseases in homeless. Ann N Y Acad Sci 2006; 1078:223–35. [DOI] [PubMed] [Google Scholar]
- 46. Badiaga S, Menard A, Tissot Dupont H et al. Prevalence of skin infections in sheltered homeless. Eur J Dermatol 2005; 15:382–6. [PubMed] [Google Scholar]
- 47. Arnaud A, Fagot-Campagna A, Reach G et al. Prevalence and characteristics of diabetes among homeless people attending shelters in Paris, France, 2006. Eur J Public Health 2010; 20:601–3. [DOI] [PubMed] [Google Scholar]
- 48. National Association of Community Health Centers . Key Health Center Data By State, 2020 (Federally-Funded Health Centers Only). Based on 2019 Uniform Data System, Bureau of Primary Health Care, HRSA, DHHS. Available at: https://www.nachc.org/wp-content/uploads/2020/09/2020-Key-Facts-by-State.pdf (last accessed 30 August 2025).
- 49. Viola KV, Tolpinrud WL, Gross CP et al. Outcomes of referral to dermatology for suspicious lesions: implications for teledermatology. Arch Dermatol 2011; 147:556–60. [DOI] [PubMed] [Google Scholar]
- 50. Omara S, Wen D, Ng B et al. Identification of incidental skin cancers among adults referred to dermatologists for suspicious skin lesions. JAMA Netw Open 2020; 3:e2030107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Kingsley-Loso JL, Grey KR, Hanson JL et al. Incidental lesions found in veterans referred to dermatology: the value of a dermatologic examination. J Am Acad Dermatol 2015; 72:651–5. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
This study is a secondary data analysis using controlled tier data from the National Institutes of Health’s All of Us Research Program. Authorized users may access this data.