Table 4. Effect of Kenya CT-OVC on sexual debut among males and females 15–25 years in 2011.

	All		Females		Males
	Unadjusted	Adjusted	Unadjusted	Adjusted	Unadjusted	Adjusted
Household in CT-OVC	0.702	0.689	0.581	0.576	0.788	0.742
95% CI	(0.555 – 0.889)	(0.531 – 0.893)	(0.393 – 0.858)	(0.383 – 0.868)	(0.585 – 1.061)	(0.531 – 1.037)
P-value	0.003	0.005	0.006	0.008	0.117	0.081
Age of head (years)		1.004		1.003		1.004
95% CI		(0.997 – 1.010)		(0.993 – 1.014)		(0.995 – 1.012)
P-value		0.306		0.535		0.383
Head is female		1.414		1.984		1.146
95% CI		(1.097 – 1.822)		(1.299 – 3.029)		(0.832 – 1.579)
P-value		0.007		0.002		0.404
Head not completed primary school		1.045		1.042		1.045
95% CI		(1.015 – 1.076)		(0.994 – 1.093)		(1.006 – 1.086)
P-value		0.003		0.085		0.023
Respondent is child of head		0.954		0.755		1.065
95% CI		(0.683 – 1.331)		(0.464 – 1.227)		(0.682 – 1.662)
P-value		0.781		0.257		0.782
Respondent is grandchild of head		1.350		0.667		1.959
95% CI		(0.931 – 1.957)		(0.379 – 1.175)		(1.191 – 3.224)
P-value		0.113		0.161		0.008
Nairobi residence		0.792		0.603		1.001
95% CI		(0.546 – 1.151)		(0.345 – 1.055)		(0.605 – 1.657)
P-value		0.222		0.076		0.997
Respondent female		0.877
95% CI		(0.693 – 1.109)
P-value		0.273
Respondent age (years)		1.265		1.260		1.277
95% CI		(1.196 – 1.338)		(1.153 – 1.377)		(1.189 – 1.371)
P-value		<0.001		<0.001		<0.001
Observations	1,433		553		880

Each column is a logistic regression model estimated on individual data and the dependent variable equals 1 if respondent has had sex and 0 otherwise. Odds ratios (OR) for each independent variable reported in corresponding row. Sample restricted to those who had not yet had sex at baseline.