Table 2. Ordinary least squares estimates of associations between SGA and age at menarche in months and between maternal characteristics and age at menarche.
Unadjusted | Adjusted for maternal demographics | Adjusted for maternal demographics, pre-pregnancy obesity, pre-pregnancy overweight, and prenatal smoking | |
---|---|---|---|
OLS coefficient | OLS coefficient | OLS coefficient | |
(95% CI) | (95% CI) | (95% CI) | |
SGA | -1.27 | -1.34 | -1.18 |
(-4.54, 2.00) | (-4.64, 1.97) | (-4.55, 2.19) | |
Race-ethnicity | |||
Non-Hispanic White | ref | ref | |
Non-Hispanic Black | -4.82 | -4.50 | |
(-7.38, -2.26) | (-7.14, -1.86) | ||
Hispanic | -6.70 | -6.66 | |
(-9.79, -3.61) | (-9.82, -3.51) | ||
Other | -2.81 | -2.84 | |
(-9.37, 3.74) | (-9.38, 3.70) | ||
Education | |||
High school graduate | -1.37 | -1.69 | |
(-3.73, 1.00) | (-4.07, 0.69) | ||
Less than high school | ref | ref | |
Nativity | |||
US born | ref | ||
Foreign born | 0.46 | 0.01 | |
(-2.82, 3.75) | (-3.29, 3.31) | ||
Age | |||
< 20 years | 0.27 | -0.66 | |
(-3.80, 4.35) | (-4.75, 3.43) | ||
20–34 | 0.72 | 0.32 | |
(-2.40, 3.85) | (-2.84, 3.49) | ||
35+ | ref | ref | |
Relationship with child’s father | |||
Married | 1.70 | 1.49 | |
(-1.23, 4.64) | (-1.42, 4.40) | ||
Cohabiting but not married | -0.47 | -0.38 | |
(-2.78, 1.83) | (-2.69, 1.93) | ||
Neither married nor cohabiting | ref | ref | |
Parity | |||
First birth | 0.09 | -0.25 | |
(-2.01, 2.19) | (-2.37, 1.87) | ||
Second of higher order birth | ref | ref | |
Health insurance for birth | |||
Medicaid | 1.49 | 1.45 | |
(-0.80, 3.79) | (-0.86, 3.75) | ||
Other | ref | ref | |
Pre-pregnancy weight | |||
Obese | -3.16 | ||
(-5.78, -0.54) | |||
Overweight | -2.90 | ||
(-5.52, -0.28) | |||
Normal or underweight | ref | ||
Prenatal smoking | |||
Any | -1.62 | ||
(-4.15, 0.91) | |||
None | ref |
Notes: OLS = Ordinary Least Squares. SGA = small-for-gestational age = <10th percentile of sex-specific birth weight for gestational age. (N = 1,003). SGA was based on birth weight and last menstrual period (LMP)-based gestational age and compared to the national U.S. sex-specific distribution of LMP-based SGA from 1999–2000. All models included a binary indicator for missing data on age in menarche in months (i.e., the respondent reported her age at menarche but not in months in addition to her years of age) and the fully adjusted model included a binary indicator for missing pre-pregnancy weight. Statistically significant estimates at p < .05 are in bold.