General Characteristics of the Sample
Among the demographic characteristics of the sample, sex differences
were compared in relation to variables with a previously established
association with high blood pressure, e.g. age, body mass, race,
education and occupation (Table 1). Percentage differences were
statistically significant (P < 0.05). The men in the worksites
screened were older (41 ± 13 versus 35 ± 12 years,
range 16-89 years), taller (176 ± 8 versus 163 ±
7cm) and heavier (80 ± 13 versus 62 ± 13kg) than
the women. The men had a larger arm circumference (30 ±
3 versus 27 ± 4cm), more education (38% male versus 30%
female college graduates), and were more likely to be white (77%
versus 58%). A1though the men were less likely to be in white-collar
jobs, when these white-collar jobs were further separated into
clerical, managerial or other, 49% of the women were in clerical
white&SHY;collar jobs, while only 19% of the men were in this
type of positions.
Prevalence of Hypertension by Sex, Age and Worksite
Table 2 compares the prevalence of hypertension across sites by
sex and age. Differences between prevalence rates were statistically
significant for males (P< 0.001). The overall prevalence of
high blood pressure among males was 26%. The site&SHY;specific
rate was highest in the two worksites with older, blue&SHY;collar
workers, typographers and skilled crafts and sanitation men, with
48 and 33% hypertension, respectively. However, among men aged
50-60 years, brokerage workers were most likely to be hypertensive
(53%). The higher crude rates for typographers thus reflected
their greater age. The surprisingly high prevalence of hypertension
in the stock&SHY;brokerage, where the high educational level suggests
a lower than average risk of hypertension, is further examined
in the multivariate analysis below, comparing other factors for
the individual subject (beyond the occupation and the work environment)
which may be related to this outcome.
Among women, the overall prevalence of hypertension was 12%. Differences
in prevalence rates for women across sites were also statistically
significant (P < 0.001), with the highest prevalence observed
in the warehouse, where women were mainly employed in time&SHY;paced
packing jobs. When the age&SHY;adjusted rates are compared, differences
across sites were significant only for women aged over 50 years
(P < 0.05). The highest rates were observed in the warehouse
and stock&SHY;brokerage (57 and 50%, respectively) for this age
group. Although the warehouse was the site where the proportion
of non&SHY;white women was highest (71%), the women at both sites
were over represented in clerical or unskilled jobs.
Multivariate analysis of blood pressure differences in 1766 men
and 1000 women with complete data, we used analysis of covariance
models, testing effects in a stepwise manner, to control simultaneously
for factors that might explain blood pressure differences between
workers. Of the variance in systolic pressure, 34% was predicted
by eight variables (Table 3). While the strongest effects on blood
pressure variation were due to age, body mass index and arm circumference
(all P~ 0.001), differences between males and females (7.2 mmHg),
worksites (9.0 mmHg), years of education completed (4.3 mmHg),
marital status (1.8mnHg) and occupational category (2.9 mmHg)
accounted for higher systolic pressures. Similar, although somewhat
weaker, results for diastolic pressure (Table 4) suggest that
after controlling for biological characteristics such as age,
body mass index and arm circumference, blood pressure levels differed
more by worksite than by any other demographic variable.
Blood Pressure and Race
Consistent with the observation of a higher prevalence of hypertension
in the USA among blacks than whites, Table 4 indicates that race
was significantly associated with diastolic blood pressure differences
(P < 0.05). However, marital status and education were not
significant in the equations for diastolic pressure, possibly
due to the relatively small sample of blacks.
Blood Pressure and Gender
To test for interactions among sex, biological, demographic and
work environment variables, separate analyses were conducted for
males and females. The results, while somewhat weaker for men,
were similar to those for the previous analysis, with worksite
accounting for the highest percentage of explained variance, after
sex was controlled (Tables 5-8). Among men, worksite, marital
status and education were related to systolic blood pressure after
controlling for biological covariates. Among women, only worksite
and occupation were related to blood pressure differences, after
control variables were entered.
This information was taken from the article: Schlussel YR, Schnall PL, Zimbler M, Warren K, Pickering TG. The Effect of Work Environments on Blood Pressure: Evidence From 7 New York Organizations, The Journal of Hypertension 8:679-685, 1990.