Descriptive Statistics for Study Sample
As expected, age and BMI were also significantly related
to case-control status in the logistic regression model. Regular
alcohol consumption was the only other variable found to have
a substantial relationship to case-control status, with an adjusted
odds ratio of 2.8 (P=.01) after controlling for other risk factors.
Cases and controls did not differ significantly on any of the
administered psychological measures after controlling for age
(Table 4).
To assess the validity of casual BP measurement for
determining case-control status and to decrease misclassification
of subjects, a further analysis was performed using the criterion
of average work-time DBP determined by the ABPM. Application of
this selection criterion with cutoff points of greater than 85,
greater than 90, and greater than 95 mm Hg applied to the cases
and 85 mm Hg or less for the controls resulted in three groups:
group 1, n=167 (67 cases and 100 controls); group 2, n=145 (45
cases and 100 controls; and group 3, n=123 (23 cases and 100 controls).
Logistical regression analyses, controlling for the same variables
as above, showed that job strain was a highly significant predictor
of case-control status, with adjusted odds ratios of 3.1 for group
1 (95% CI, 1.2 to 8.0), 3.6 for group 2 (95% CI, 1.2 to 10.7,
and 24.4 for group 3 (95% CI, 3.6 to 167.0). Despite the sharp
reductions in sample size that result from adding an additional
selection criterion based on work-time ambulatory BP, the strength
and statistical significance of the relationship of job strain
to case-control status are clearly greater when we use this more
stringent definition of cases and controls.
For more information regarding this site, e-mail us
at: cse@workhealth.org