Table 4.

Adjusted odds ratios from logistic models predicting overweight/obese as a function of ethnicity–nativity and duration of stay in the US, Mexican-origin and white respondents ages 4–24, CHIS (2005, 2007, 2009, N = 25,487).

Outcome: overweight/obese	N	Ethnicity/nativity
		Mexican-origin, generation				US-born non-Hispanic white AOR (se)
		1st in US <5 years 1st in US ≥5 years AOR (se)		2nd AOR (se)	3+ (ref. group) AOR (se)
Full sample	25,487	1.12 (0.17)	1.09 (0.12)	1.18 (0.12)*	1.00 (–)	0.58 (0.04)**
Child (ages 4–11)
Female	5678	0.77 (0.40)	1.18 (0.52)	1.01 (0.17)	1.00 (–)	0.50 (0.07)**
Male	6065	1.61 (0.73)	1.09 (0.40)	1.12 (0.23)	1.00 (–)	0.50 (0.07)**
Adolescent (ages 12–17)
Female	4055	0.65 (0.27)	0.81 (0.24)	1.01 (0.21)	1.00 (–)	0.36 (0.06)**
Male	4222	1.50 (0.56)	1.01 (0.31)	1.38 (0.30)	1.00 (–)	0.72 (0.11)*
Adult (ages 18–24)
Female	2672	1.49 (0.46)	0.90 (0.21)	1.14 (0.24)	1.00 (–)	0.71 (0.13)
Male	2795	1.26 (0.37)	1.75 (0.37) **	1.72 (0.38)*	1.00 (–)	0.67 (0.12)*

Significantly different from 3+ generation Mexican-origin at

^*p < 0.05,

^**p < 0.01.

CHIS: California Health Interview Survey. Each table row reports adjusted odds ratios from one logistic model. All models control for age in years. Model on full sample also controls for sex. Models are weighted for the CHIS sampling scheme. Reported standard errors (in parentheses) are the standard errors on the log odds multiplied by the odds ratio (exponentiated log odds). Standard errors are computed using jackknife resampling.