Housing Instability and Health Care Engagement Among People With CKD.

To the Editor:

Housing instability affects more than 38.1 million American households and is characterized by high housing costs or overcrowded living conditions., Engagement with nephrology care is of interest in this population because housing instability has been associated with risk factors for chronic kidney disease (CKD), and homeless persons with CKD are more likely to progress to end-stage kidney disease than people with stable housing. In the general population, housing instability is associated with not having a usual source of care and postponing medical care., The relation of housing instability to health care use among people with CKD is unknown. We sought to examine whether housing instability was associated with postponing medical care among persons with CKD and hypothesized that those experiencing housing instability would be more likely to report postponing medical care that they believed to be necessary.

We used data from the Healthy Aging in Neighborhoods of Diversity Across the Life Span (HANDLS) study (Baltimore, MD). HANDLS is a population-based cohort study examining the influence and interaction of race and socioeconomic status on the development of health disparities among minorities and lower-socioeconomic-status groups. The cohort included 3,720 black and white community-dwelling individuals between the ages of 30 and 64 years who were sampled from 13 socioeconomically diverse neighborhoods. Participants were enrolled between August 2004 and November 2008. Each participant provided written informed consent, and the National Institute of Environmental Health Sciences, National Institutes of Health approved the study protocol (09-AG-N248).

We performed a cross-sectional analysis evaluating the association between housing instability and postponement of medical care at HANDLS visit 4 (September 2013 to September 2017). Individuals were included if they attended visit 4 and had prevalent CKD (estimated glomerular filtration rate < 60 mL/min/1.73 m2 and/or urinary albumin-creatinine ratio ≥ 30 mg/g; N = 558). Individuals were excluded if they had missing data (N = 203), leaving a study population of N = 355. Housing instability was defined as a negative response to the question “Are you able to afford a suitable home for you and your family?” or answering a positive response to “Have you had difficulties making rent or mortgage payments?” Poor health care engagement was defined as a positive response to the question “Have you postponed needed health care since the last HANDLS exam?” at study visit 4. All covariables of interest were ascertained at visit 4.

We compared participant characteristics using χ2 tests for categorical variables and t tests for continuous variables. We estimated the prevalence of postponing needed health care overall and among those with and without housing instability. We used multivariable log binomial and Poisson regression with robust estimate of variance clustered on neighborhood to quantify associations between housing instability and self-report of postponing medical care that was believed to be needed and adjusted for demographics (age, race, sex, and poverty status), clinical variables (estimated glomerular filtration rate, log-transformed urinary albumin-creatinine ratio, hypertension status, and diabetes status), health insurance status, CKD awareness, food insecurity, and education level. We evaluated for potential effect modification by race by including an interaction term for race × housing instability. As sensitivity analysis, we evaluated associations for HANDLS participants without CKD and tested for effect modification by CKD status.

Among 355 HANDLS participants with CKD, 135 (38%) reported housing instability. Individuals with housing instability were younger (mean [standard deviation] age, 57.8 [9.1] vs 61.1 [8.3] years), were less likely to have a high school degree, and were more likely to be current smokers and report food insecurity than stably housed persons (P < 0.05 for all; Table 1).

Table 1

Baseline Characteristics According to Housing Status

Variable	No Housing Instability (N = 220)	Housing Instability (N = 135)	P
Age, y	61.6 (8.3)	57.8 (9.1)	<0.001
Male sex at birth	85 (38.6%)	66 (48.9%)	0.06
Black race	148 (67.3%)	97 (71.9%)	0.37
Annual income < 125% federal poverty level	99 (45.0%)	59 (43.7%)	0.81
eGFR, mL/min/1.73 m2	70.1 (27.5)	77.3 (28.3)	0.02
UACR, mg/g	53 [14-146.5]	64 [31-171]	0.21
Albumin, g/dL	4.19 (0.36)	4.14 (0.37)	0.20
Systolic blood pressure, mm Hg	122.6 (23.0)	122.7 (18.5)	0.98
Diabetes	108 (49.1%)	67 (49.6%)	0.92
Coronary artery disease	22 (10%)	21 (15.6%)	0.12
Chronic obstructive pulmonary disease	36 (16.4%)	16 (11.9%)	0.24
History of stroke	19 (8.6%)	9 (6.7%)	0.50
Current smokera	74 (37.6%)	70 (56.0%)	<0.001
Food insecurity	33 (15.0%)	67 (49.6%)	<0.001
Enough money for medical care	163 (74.1%)	56 (41.5%)	<0.001
≥High school degree	159 (72.3%)	82 (60.7%)	0.02
No health insurance	9 (4.1%)	9 (6.7%)	0.28
Aware of CKD	47 (21.4%)	28 (20.7%)	0.89

Note: Values for categorical variables given as number (percent); values for continuous variables are given as mean (standard deviation) or median [interquartile range].

Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; UACR, urinary albumin-creatinine ratio.

Smoking status was missing for 23 participants in the no housing instability group (total N = 197) and 95 participants in the housing instability group (total N = 125).

Overall, 85 (23.9%) participants reported postponement of medical care. Housing instability was associated with increased risk for postponing medical care in unadjusted (unadjusted incidence rate ratio [IRR], 1.92; 95% confidence interval [CI], 1.44-2.56) and adjusted models (IRR, 1.59; 95% CI, 1.20-2.10; Tables 2 and S1). The relationship between housing instability and postponement of medical care was not modified by race (P for interaction = 0.30). Associations were stronger for people with CKD compared with participants without CKD, although not statistically significant (P interaction = 0.21; Table S2).

Table 2

Association Between Housing Instability and Postponement of Medical Care Among People With CKD

Analysis	Incidence Rate Ratio (95% CI)
Unadjusted	1.92 (1.44-2.56)
Model 1	1.73 (1.34-2.23)
Model 2	1.72 (1.36-2.18)
Model 3	1.60 (1.21-2.12)

Note: Model 1 adjusted for demographics (age, race, sex, and poverty level). Model 2 adjusted for demographics and clinical variables (baseline estimated glomerular filtration rate, log-transformed urinary albumin-creatinine ratio, blood pressure, and diabetes). Model 3 adjusted for demographics, clinical variables, health insurance, CKD awareness, food insecurity, and education level.

Abbreviations: CKD, chronic kidney disease; CI, confidence interval.

Our study had limitations. We lacked data for more traditional indicators of housing instability, such as rent to income ratio or measure of persons per room in the home. Therefore, our study may not have captured all participants experiencing housing instability. Our sample was moderately sized, and findings may not be generalizable to nonurban populations. Due to the cross-sectional design, our findings cannot be used to draw conclusions about temporality or causality.

In this cross-sectional analysis of 355 urban-dwelling individuals with CKD, we found that those experiencing housing instability were more likely to report postponing needed medical care. Postponing medical care could increase risk for poor clinical outcomes in this vulnerable population. Risk reduction efforts specifically targeting persons with CKD who are experiencing housing instability should be considered.