Table 4.
PCS | MCS | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Intersectionality: race * sex | ||||||||||||
Base model | Race * sex interactions | Base model | Race * sex interactions | |||||||||
Variables | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value |
Constant (White men) | 50.62 | (50.08,51.16) | <0.0005 | 50.61 | (50.01,51.22) | <0.0005 | 53.95 | (53.47,54.43) | <0.0005 | 54.01 | (53.51,54.50) | <0.0005 |
Education | 0.91 | (0.79,1.04) | <0.0005 | 0.92 | (0.73,1.10) | <0.0005 | 0.73 | (0.44,1.02) | <0.0005 | 0.44 | (−0.03,0.90) | 0.066 |
White women | −1.73 | (−2.39,−1.08) | <0.0005 | −1.74 | (−2.58,−0.89) | <0.0005 | −2.54 | (−3.16,−1.91) | <0.0005 | −2.68 | (−3.35,−2.01) | <0.0005 |
Black men | 0.09 | (−0.80,0.97) | 0.848 | 0.22 | (−0.76,1.19) | 0.661 | −0.11 | (−0.90,0.67) | 0.773 | −0.18 | (−0.96,0.61) | 0.658 |
Black women | −2.72 | (−3.62,−1.83) | <0.0005 | −2.77 | (−3.81,−1.73) | <0.0005 | −1.89 | (−2.69,−1.10) | <0.0005 | −2.03 | (−2.84,−1.22) | <0.0005 |
Hispanic / Latino men | 0.46 | (−0.70,1.63) | 0.437 | 0.48 | (−0.78,1.74) | 0.456 | 0.63 | (−0.41,1.68) | 0.234 | 0.58 | (−0.47,1.63) | 0.279 |
Hispanic / Latino women | −0.50 | (−1.89,0.89) | 0.484 | −0.38 | (−1.84,1.08) | 0.612 | −0.86 | (−2.00,0.27) | 0.136 | −0.97 | (−2.10,0.17) | 0.095 |
White women * education | 0.00 | (−0.29,0.30) | 0.990 | 0.65 | (−0.06,1.37) | 0.075 | ||||||
Black men * education | −0.30 | (−0.68,0.08) | 0.119 | 0.22 | (−0.62,1.06) | 0.608 | ||||||
Black women * education | 0.06 | (−0.33,0.45) | 0.763 | 0.91 | (0.19,1.64) | 0.014 | ||||||
Hispanic men * education | 0.00 | (−0.44,0.43) | 0.983 | 0.36 | (−0.42,1.14) | 0.364 | ||||||
Hispanic / Latino women * education | −0.29 | (−0.77,0.20) | 0.249 | −0.13 | (−0.99,0.74) | 0.775 | ||||||
Intersectionality: cSES * race | ||||||||||||
Base model | cSES * race interactions | Base model | cSES * race interactions | |||||||||
Variables | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value |
Constant (White, high cSES) | 50.65 | (50.13, 51.18) | < 0.0005 | 50.77 | (50.20, 51.33) | < 0.0005 | 53.76 | (53.26, 54.26) | < 0.0005 | 53.90 | (53.37, 54.43) | < 0.0005 |
Education | 0.91 | (0.78, 1.04) | < 0.0005 | 0.83 | (0.67, 0,99) | < 0.0005 | 0.26 | (0.14, 0.39) | < 0.0005 | 0.19 | (0.03, 0.35) | 0.023 |
White, low cSES | −1.64 | (−2.51, −0.78) | < 0.0005 | −1.75 | (−2.65, −0.84) | < 0.0005 | −0.78 | (−1.59, 0.03) | 0.058 | −0.90 | (−1.74, −0.06) | 0.035 |
Black, high cSES | −0.82 | (−1.63, −0.00) | 0.049 | −0.64 | (−1.66, 0.38) | 0.216 | 0.01 | (−0.74, 0.76) | 0.983 | −0.26 | (−1.17, 0.65) | 0.573 |
Black, low cSES | −1.71 | (−2.63, −0.79) | < 0.0005 | −1.88 | (−2.85, −0.92) | < 0.0005 | −0.08 | (−0.85, 0.70) | 0.848 | −0.27 | (−1.09, 0.55) | 0.518 |
Hispanic / Latino, high cSES | 0.84 | (−0.62, 2.30) | 0.259 | 1.15 | (−0.85, 2.15) | 0.261 | 1.68 | (0.51, 2.85) | 0.005 | 1.54 | (0.05, 3.04) | 0.043 |
Hispanic / Latino, low cSES | −0.72 | (−1.83, 0.40) | 0.208 | −0.86 | (−1.99, 0.27) | 0.136 | 0.27 | (−0.67, 1.21) | 0.575 | 0.15 | (−0.81, 1.10) | 0.762 |
White, low cSES * education | 0.38 | (−0.08, 0.83) | 0.105 | 0.21 | (−0.23, 0.65) | 0.356 | ||||||
Black, high cSES * education | −0.21 | (−0.58, 0.15) | 0.254 | 0.19 | (−0.18, 0.57) | 0.309 | ||||||
Black, low cSES * education | 0.19 | (−0.21, 0.58) | 0.354 | 0.61 | (0.21, 1.00) | 0.002 | ||||||
Hispanic / Latino, high cSES *education | −0.22 | (−0.74, 0.30) | 0.415 | 0.08 | (−0.35, 0.51) | 0.725 | ||||||
Hispanic / Latino, low cSES * education | 0.01 | (−0.49, 0.50) | 0.978 | 0.33 | (−0.09, 0.76) | 0.127 | ||||||
Intersectionality: cSES * sex | ||||||||||||
Base model | cSES * sex interactions | Base model | cSES * sex interactions | |||||||||
Beta | (95% CI) | p-value | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value | Beta | (95% CI) | p-value | |
Constant (high cSES men) | 50.68 | (50.16, 51.21) | < 0.0005 | 50.87 | (50.27, 51.46) | < 0.0005 | 53.92 | (53.46, 54.38) | < 0.0005 | 54.19 | (53.70, 54.68) | < 0.0005 |
Education | 0.91 | (0.78, 1.04) | < 0.0005 | 0.80 | (0.64, 0.96) | < 0.0005 | 0.73 | (0.44, 1.02) | < 0.0005 | 0.07 | (−0.39, 0.54) | 0.760 |
High cSES women | −1.85 | (−2.47, −1.22) | < 0.0005 | −1.87 | (−2.76, −0.99) | < 0.0005 | −2.49 | (−3.09, −1.89) | < 0.0005 | −3.06 | (−3.83, −2.30) | < 0.0005 |
Low cSES men | −1.78 | (−2.66, −0.90) | < 0.0005 | −1.92 | (−2.85, −0.99) | < 0.0005 | −0.77 | (−1.56, 0.02) | 0.056 | −1.02 | (−1.80, −0.23) | 0.012 |
Low cSES women | −3.21 | (−4.10, −2.31) | < 0.0005 | −3.39 | (−4.34, −2.44) | < 0.0005 | −2.96 | (−3.78, −2.14) | < 0.0005 | −3.19 | (−4.02, −2.36) | < 0.0005 |
High cSES women * education | 0.02 | (−0.26, 0.30) | 0.886 | 1.32 | (0.46, 2.18) | 0.003 | ||||||
Low cSES men * education | 0.38 | (−0.05, 0.82) | 0.085 | 0.68 | (−0.04, 1.40) | 0.064 | ||||||
Low cSES women * education | 0.41 | (−0.06, 0.89) | 0.086 | 0.75 | (0.01, 1.49) | 0.048 |
PCS is physical health component summary score; MCS is mental health component summary score.
Education is centered at 12 years, and birth year is centered at 1960 so the constant is interpretable.
For PCS, education is coded linearly as years of education; a 1-unit increase in education is a 1-year increase in schooling.
For MCS, education is coded linearly as years of education until 13 years, and flat thereafter (i.e. everything after 13 years of schooling is recoded as 13 years because found this specification was the best fit for the data). A 1unit increase in education is a 1-year increase in schooling until 13 years.
Models adjusted for all effect modifies and confounders; we modeled main effects and interactions for Other race men and Other race women but did not display these results due to ambiguity in interpretation.
The reference group is White men born in the U.S., outside the South, who lived in non-rural areas at age 14 years and whose mothers completed 12 or more years of schooling; because each analytic model varies in the interaction terms included, the interpretation of the reference group similarly varies across models.
Interaction term coefficients indicate the additional difference in PCS / MCS associated with each year of education; positive interaction terms indicate the demographic group benefited more from each year of education than the reference group, while negative interaction terms indicate the demographic group benefited less from each year of education than the reference group. To calculate the change in PCS / MCS for a one-year increase in education for a specific demographic group, sum the coefficient for the main effect for education with the coefficient for the interaction term.