. Author manuscript; available in PMC: 2010 Apr 19.

Published in final edited form as: J Aging Health. 2009 Jun 10;21(5):759–775. doi: 10.1177/0898264309333312

Table 3.

Multiple Logistical Regression for Predictive Value of Specific Health Behaviors on CVD Diagnosis Controlling for Age and Gender (Max-Rescaled R² = .2868)

	CVD Diagnosis OR^a (95% CI)
Independent Variables	n = 525
Age group
Young (25–44)	Reference
Middle-aged (45–64)	4.355 (2.62–7.25)^*
Young-old (65–74)	11.358 (6.66–19.36)^*
Old-old (75–91)	28.394 (15.33–52.58)^*
Gender
Female	Reference
Male	1.798 (1.34–2.42)^*
Health behavior^b
Obese	2.041 (1.39–3.00)^*
Risky medical	0.554 (0.39–0.78)^*

Note: CVD = cardiovascular disease; OR = odds ratio; CI = confidence interval.

Indicates controlling for all variables in the model.

Indicates health behavior reference variable is nonrisky status.

p ≤ .05.