The Humanistic and Economic Burden Associated with Anxiety and Depression among Adults with Comorbid Diabetes and Hypertension

Table 4

Unadjusted and adjusted association between depression and anxiety categories and high out-of-pocket (OOP) spending burden among adults with comorbid diabetes and hypertension, using pooled data from the 2013 and 2015 Medical Expenditure Panel Survey.

Bivariate association between depression and anxiety categories and high OOP spending burden
	High burden^†		Not high burden
		Wt row%		Wt row%	value	Sig.
Depression and anxiety	90	34.0	219	66.0	<0.001
Depression only	158	28.2	403	71.8
Anxiety only	96	23.5	270	76.5
No depression/anxiety	650	18.3	2674	81.7
Logistic regressions on high OOP spending burden^†
	Unadjusted model			Adjusted model^‡
	OR	95% CI	Sig.	AOR	95% CI	Sig.
Depression and anxiety	2.26	(1.66–3.07)		1.55	(1.06–2.25)
Depression only	1.72	(1.28–2.32)		1.25	(0.91–1.70)
Anxiety only	1.34	(1.04–1.74)		1.03	(0.77–1.38)
No depression/anxiety	(reference group)

Note: the analytic sample consisted of adults (>18 years) with comorbid diabetes and hypertension who were alive and had positive expenditures in the observation year (2013/2015). Wt row%: weighted row percentages; Sig.: statistical significance level; OR: odds ratio; AOR: adjusted odds ratio; 95% CI: 95% confidence interval. ^†OOP spending burden was measured by the percentage of household income spent on healthcare. It was calculated by dividing OOP spending by household income. High OOP spending burden was defined as 10% or more. ^‡Covariates adjusted in the logistic regression included predisposing factors (sex, age groups, and race/ethnicity), enabling factors (marital status, education level, poverty status, health insurance coverage, and prescription drug insurance coverage), enabling factor (presence/absence of other chronic conditions, pain interference), personal health practice (obesity, smoking status, and exercise level), and external environment (region). ; ; .