Derivation of Study Data
Once the WIC cases and matched controls were selected,
all data from their birth certificates were extracted for
analysis. The Massachusetts birth certificate provides data
on maternal demographic characteristics, prenatal care, and
pregnancy outcomes.6 The State Death Registry file provides
information on all neonatal deaths between birth and
28 days.
Data Analysis
The demographic characteristics of WIC participants
were contrasted with the characteristics of all pregnant
women residing in the same catchment area and statewide in
1978. WIC participants were then compared directly to their
matched non-WIC controls on the birth outcome measures.
Differences were statistically examined by use of paired Ttest
comparisons for continuous data items and by McNemar
Chi-square comparisons for ordinal data items. Pairwise
deletions were used for any subject pair having missing data.
The women in the WIC sample and their matched controls
were then stratified into a number of subpopulations for
separate analyses of birth outcome differences. These subgroups
were defined on the basis of demographic characteristics
or duration of WIC participation and are not statistically
independent of each other.
Results
WIC Population Characteristics-The comparison of
selected demographic characteristics (Table 1) suggests that
the WIC population comes from demographic groups at
higher risk for poor pregnancy outcomes. The WIC prenatal
population is younger and less educated, contains more
unmarried and minority women, and has a larger number of
high parity births compared to all women who reside in the
same WIC catchment area or statewide and gave birth in
1978. The Massachusetts WIC program sites in 1978 were
located in more disadvantaged areas of the state and, within
the program, the higher risk subpopulations were proportionally
more heavily represented. Almost 18 per cent of all
women who gave birth in 1978 in WIC catchment areas were
WIC prenatal participants.
Overall WIC and Control Group Comparison-T