Gender and Age Differences in Lifestyle Factors Related to Hypertension in Middle-Aged Civil Service Employees

The aim of this study is to identify lifestyle factors related to hypertension in man and woman workers, and to investigate age and gender differences in the relationships of the factors. From 6,000 civil service employees (4,937 men and 1,063 women) aged 40-69 years, information on lifestyle-related factors such as stress, exercise habits, preference for salty taste, alcohol drinking and smoking habits, and body mass index, as well as age and family history of hypertension was obtained through self-administered questionnaires in 1997. Hypertension was defined as either a systolic blood pressure ≧ 140mmHg, a diastolic blood pressure ≧ 90 mmHg, or undergoing treatment for hypertension, and was present by 37.0% in men and 19.6% in women. Only body mass index was a significant lifestyle-related risk factor common to both genders with an odds ratio and its 95% confidence interval in parentheses of 2.2 (2.0 - 2.5) for men and 3.2 (2.3 - 4.6) for women. Men and women who preferred salty taste showed multivariate adjusted odds ratios of 0.9 (0.8 -1.1) and 1.5 (1.1 - 2.2) for hypertension, respectively. In the stratified subanalysis, women aged 50 years and over had a significant odds ratio of 2.7 (1.5 - 4.9), whereas women aged 40-49 years and men of all age classes failed to show significant relationships. Salt intake was suggested to be a key factor for hypertension particularly for women after menopause.

Among lifestyle-related diseases, hypertension is a major risk factor for cardiovascular diseases in both men and women.' Because hypertension has a high detection rate in annual health check-ups, it has become an important illness from the standpoint of health care of employees. Epidimemiology studies performed in Western countries as well as in Japan have clearly shown that there was a gender difference in the morbidity of hypertension.'' Some studies performed in Western countries have also shown further that the morbidity of hypertension in women increases after menopause.',' This finding attracts attention from the viewpoint of health guidance for hypertension in middle-aged women. In occupational health field in Japan, however, the health strate-gies for women have been set to protect the motherhood, and they did not pay enough attention to life-style related diseases.' Therefore, we started collection of annual health check-up data of middle-aged civil service employees and made a questionnaire survey on their lifestyle in 1997.
In the present study, we did a cross-sectional analysis of the Subjects were classified into two groups according to their response to the question, 'How much daily stress do you feel?' on the questionnaire. Subjects who answered 'very much' or 'fairly much' were combined to form the 'high stress' group, while those who answered 'the usual' or 'little' were combined into the 'low stress' group. Subjects were also grouped according to their answer to the question, 'Do you normally exercise or engage in sports other than work?' Those who responded that they exercised one or more times a week were placed in an 'exercise habit' group, and those who exercised one to three times per month or almost never in a 'no exercise habit' group.
As to a preference for saltiness, subjects were classified into three groups based on their response to the question, 'Do you like food strongly seasoned with salt, soy sauce, and miso (soybean paste)?' Response choices were 'prefer salty taste and often eat foods so seasoned', 'prefer salty taste but abstain from foods so seasoned', 'prefer bland taste', or 'no special preference'. Those who chose either one of the first two answers were combined into a 'prefer salty taste' group, and this group was compared with the other two groups.
Those responding that they consumed alcohol once or more a week were classified into an 'alcohol habit' group, while those who consumed alcohol 'almost never' or 'not at all' were listed as a 'no alcohol habit' group . Those who responded that they smoked were placed in a 'smoking habit' group, and those who had quit smoking or never smoked were combined into a 'no smoking habit' group.
BMI was calculated from the height and weight measurements in the health check-up using the formula 'weight (kg)/square of height (m)', and was used as the indicator of obesity. Subjects were classified into two groups according to BMI with a cutoff value of 24.0 (kg/m2), the integer closest to the former standard9 for being overweight of the Japan Society for the Study of Obesity.
In addition to these lifestyle habits, a family history of hypertension and age were included as factors for analysis. The family history was judged based on information on history of hypertension in both parents and siblings. Subjects were grouped by age into four classes of 40-44, 45-49, 50-54, and 55 years and older.

Statistical analysis
The relationship between hypertension and each factor was investigated using a chi-square test. The Cochran-Armitage method was used to test the trend of proportion. In a comparison among the three groups, the significance level was corrected using Bonferroni's methods. To investigate items having indepen-dent relations with hypertension even after adjusting for confounding among factors, a multivariate logistic regression analysis was employed separately for men and women, with lifestyle factors, age class, and family history of hypertension as independent variables. For the preference for saltiness and age classes, dummy variables were assigned by taking 'prefer bland taste' and '40-44' as the respective references. The strength of the relationship between the dependent variable and each of the independent variables was expressed by the odds ratio (OR) and its 95% confidence interval (CI). The statistical analysis package SFSS(1) 10.OJ for Windows was used for statistical processing. The significance level was set at p<0.05. Table 1 shows the distribution of subjects with hypertension by gender and age. The proportion of hypertensives was significantly lower in women than in men overall (19.6% vs 37.0%, p<0.05), as well as for every age class. The proportion increased significantly (p<0.001) with the advance of age in both men and women.  ***: p<0 .001 CI: confidence interval *: Confounding effects between the factor in question and all other factors on the odds ratio were adjusted simultaneously through a multivariate logistic regression analysis. Subjects with unknown response were excluded. The result of the Cochran-Armitage trend test showed a significant increase in the proportional rate of hypertensives with advancing age in men preferring a salty taste (p<0.001), men preferring a bland taste (p<0.001), women preferring a salty taste (p<0.05), and women preferring a bland taste (p<0.001). In women, the proportion of hypertensives among those who preferred a salty taste increased remarkably after the age of 50, and was significantly higher than that of hypertensives among those who preferred a bland taste.    Furthermore, it should have been necessary to measure blood pressure without distinction between normotensives and hypertensives. However, for the latter, the measurement was done twice and lower value was adopted. It is probable that this measurement bias would have yielded weaker relationship between lifestyle factors and hypertension than real one.

RESULTS
The present study has some limitations. We could not obtain information on actual salt intake or salt excretion in the urine, or information on menopause from individual subjects of this study. This renders our interpretation on the relationship among salt intake, menopause, and blood pressure speculative, although it is consistent with past experimental, epidemiologic and clinical findings. Furthermore, since this analysis was cross-sectional and the results do not necessarily explain the cause and effect relationships between the various lifestyle factors and hypertension. However, presence of gender and age-specific factors related to hypertension suggests a necessity for further study to look for efficient preventive approach.