Table 2

Contemporary and lagged pooled, cross sectional time series analyses of predicted low birth weight births as percentage of all births, US states, 1985–1995

ModelSame year†One year lag†Three year lag†Five year lag†
*0.01<p<0.05; **0.001<p<0.01; ***0.0001<p<0.001; ****p<0.0001. †Estimated coefficient and standard errors (in parentheses) were provided from mixed models. Time lagged models predict contemporary dependent variables using one, three, or five year lagged covariates. ‡Primary care physicians per 10 000 population. §Percentage state population that is African-American. ¶Percentage state population in metropolitan areas. ††Percentage of state population unemployed. ‡‡Percentage of state population with high school education or above. §§Δdf, difference of degree of freedom for two models (1 v 2, 2 v 3); ¶¶Δchsq, difference of χ2 for two models (1 v 2, 2 v 3); †††LRT (likelihood ratio test), (Δ of χ2 for two models)/(Δ of df for two models). Sources, see references.22–25,27
Gini coefficient18.3****18.2****3.0**18.4****17.8****2.118.8****17.4****2.218.2****16.9****2.7*
Primary care supply‡−0.4****−0.2****−0.4****−0.2****−0.4****−0.2****−0.5****−0.2***
% African-American§8.9****9.1****9.4****9.1****
% Metro¶−0.01****−0.01****−0.01***−0.01**