Half of the sample were women (N = 752, 50%). The median age was 49 (interquartile range 28) years; 1% had an academic degree, 51% had a university degree, 28% were college graduates, 19% had only high school and the remaining 1% only had secondary school education (Table 1). Only 2% had a household income exceeding 501,000 tenge (1290 United States dollars) a month. Forty-one percent of subjects never smoked cigarettes, 28% smoked in the past and 31% were current smokers. Sex affected almost all included sociodemographic and lifestyle determinants. Thus, more women had higher education, smaller income, demonstrated smaller prevalence of all studied tobacco products and more women were never alcohol users. The least median score of eight studied HRQL domains was 38.4 (GH), the highest—53.4 (BP) (Table 2).
In the univariate analysis, both physical and mental components were associated with age and sex (Table 3). PCS had more advanced decline with age compared with mental component. Women had significantly worse HRQL in both domains. College of university education did not affect any HRQL component, whereas monthly income had a positive association with both physical and mental scores, but the effect was greater for PCS. Subjects smoking cigarettes, electronic cigarettes or using waterpipe had a significantly greater PCS score, apparently because they were yet healthy enough to smoke. Cigarette smoking, but not waterpipe use of electronic cigarette smoking, was also associated with mental component. Regular physical activity positively affected both HRQL domains. Alcohol use was associated with higher physical component of HRQL, but not with mental.
We excluded occupations with the fraction 1% or below in the overall employment profile because of low power. These excluded occupations were expeditor, plumber, fueler, musician, filmmaker, controller, coddler, carpenter, pilot, pressman, pharmacist, geologist, turner, architect, interpreter, journalist, baker, gardener, psychologist, librarian, flight attendant, metallurgist, miner and mapmaker. A number of occupations were associated with the physical component of HRQL, but none of them retained significant association when adjusted for socioeconomic determinants or lifestyle (Model 1 in Table 4), indicating that not the occupation, but the underlying socioeconomic determinants and lifestyle predicted HRQL. Furthermore, such occupations as driver (N = 104, 6.9%), civil servant (N = 101, 6.7%), construction (N = 96, 6.4%), nurse (N = 82, 5.5%), businessman (N = 78, 5.2%), work in catering (N = 76, 5.1%), security officer (N = 58, 3.9%), foreman (N = 56, 3.7%), medical doctor (N = 41, 2.7%), military officer (N = 36, 2.4%), tailor (N = 33, 2.2%), mechanic (N = 31, 2.1%), police officer (N = 29, 1.9%), electrician (N = 27, 1.8%), machinist (N = 26, 1.7%), metal worker (N = 26, 1.7%), agricultural worker (N = 24, 1.6%), technologist (N = 22, 1.5%), artist (N = 17, 1.1%), hairdresser (N = 16, 1.1%), fitter (N = 16, 1.1%), lawyer (N = 17, 1.1%) or fitness trainer (N = 15, 1.0%), were not associated with either PCS and or MCS even in the crude models. In such an adjusted model, only age, income and physical activity could determine PCS. Of note, age and income had a negative association, whereas physical activity shower positive association. Of all included occupations, only lab technicians had some, but yet non-significant (p = 0.055) negative association with PCS in the models adjusted for socioeconomic determinants and lifestyle.
In a univariate analysis, such occupations as economist, manager, kindergarten teacher, sales person, welder and secretary were associated with poorer mental component of HRQL, but working as an IT specialist showed an opposite association. When adjusted for age, sex, monthly income, cigarette smoking and physical activity, only managers, welders and secretaries showed significantly worse MCS, more pronounced for welders and secretaries (Model 2 in Table 4). In a fully adjusted model, when significantly associated with HRQL occupations (manager, welder and secretary) were adjusted not only for age, sex, income and physical activity, but also for each other, they all remained significantly associated with MCS, assuming that such association was independent of each included occupation. Thus, in such fully adjusted model, ever being a manager reduced the mental component of HRQL by 1.63 (95% CI − 2.92; − 0.34) points; a welder—by 5.11 (95% CI − 8.77; − 1.46) and a secretary—by 5.06 (95% CI − 8.56; − 1.56) points. With regard to specific exposures, in an adjusted for sociodemographic predictors model, exposure to irritant gases for more than a year negatively affected PCS (beta − 3.20 (95% CI − 5.99; − 0.42)), whereas exposure to grain or flour in the workplace was associated with poorer MCS (beta − 3.51 (95% CI − 6.72; − 0.29)).