Table 2.

Odds ratios and predicted prevalences (in selected subgroups) of excessive gestational weight gain calculated from multivariable logistic regression model*

	Pre-pregnancy weight status perception
	Normal weight women		Overweight/obese women
	Accurate assessors n = 898	Overassessors n = 131	Accurate assessors n = 438	Underassessors n = 70
Multivariable odds ratios (95% CI)	1.0 (Referent)	2.0 (1.3, 3.0)	2.9 (2.2, 3.9)	7.6 (3.4, 17.0)
Participant characteristics modeled†	Predicted prevalence of excessive gestational weight gain (prevalence ratio) ‡
Aged 23 y, non-white, parous, lower income, less education§	37%	54% (1.5)	64% (1.7)	82% (2.2)
Aged 23 y, white, parous, lower income, less education§	45%	62% (1.4)	70% (1.6)	86% (1.9)
Aged 32 y, white, nulliparous, higher income, well-educated ||	49%	66% (1.3)	74% (1.5)	88% (1.8)

* Data from 1537 mothers participating in Project Viva.

† All groups were married, employed, non-smokers, and had mean pre-pregnancy BMI and gestation length.

‡ Prevalence ratios compare predicted prevalence in each category with normal weight accurate assessors who had the same participant characteristics.

§Household income less than or equal to $40,000 per year, some college or less.

|| Household income more than $40,000 per year, college degree.