Table 2.

Crude and adjusted models for depression among Tushirikiane participants, Kampala, Uganda, 2020 (n = 367)

	Crude Model*		Adjusted Model†
	OR (95% CI)	P-value‡	aOR (95% CI)	P-value‡
Time Period		.583		.443
Before COVID-19 declaration	ref		ref
After COVID-19 declaration	1.08 (0.82, 1.44)		1.16 (0.80, 1.67)
Demographic Factors
Age, y	1.19 (1.10, 1.30)	<.001	1.10 (1.00, 1.21)	.047
Gender		.350		.155
Man	ref		ref
Woman	1.21 (0.81, 1.82)		1.37 (0.89, 2.12)
Highest Level of Education		.157
Less than secondary	ref
Some secondary	1.62 (0.96, 2.75)		—–
Secondary +	1.18 (0.70, 2.03)
Length of time in Uganda		<.001		.009
<1 y	ref		ref
1–5 y	0.27 (0.09, 0.80)		0.55 (0.21, 1.41)
6–10 y	0.64 (0.21, 1.92)		1.18 (0.44, 3.16)
>10 y	0.51 (0.15, 1.69)		0.93 (0.31, 2.81)
Relationship Status		.632
No current partner	ref
Dating one partner and/or married	0.97 (0.62, 1.50)		—–
Casual dating and/or multiple partners	0.73 (0.38, 1.40)
Ecosocial Factors
Food Insecure		.003		.048
Never	ref		ref
Sometimes	1.44 (0.95, 2.19)		1.31 (0.82, 2.10)
Always	3.36 (1.69, 6.69)		2.54 (1.21, 5.33)
Experienced any physical and/or sexual IPV		.003		.034
No	ref		ref
Yes	2.66 (1.40, 5.07)		2.53 (1.07, 5.96)
Social Provisions Scale	0.83 (0.80, 0.87)	<.001	0.85 (0.81, 0.89)	<.001

Note: OR = odds ratio; aOR = adjusted odds ratio; CI = confidence interval; IPV = intimate partner violence.

^⁎Crude model conducted using general estimating equation logistic regression with robust standard errors to account for clustering controlling for settlement as an a priori covariate.

^†Adjusted model conducted using general estimating equation logistic regression with robust standard errors to account for clustering, controlling for settlement, time period, gender, and all significant factors from crude models.

^‡P-values calculated from Wald test.