parts producer. Only full-time employees were invited to participate in the study. At four study sites, all full-time employees were invited. At three other sites, full-time employees who participated in health checkups within a certain period were invited. At one site, full-time employed men who had participated in health checkups were invited. At another site, all supervisors and managers were invited. A questionnaire was distributed by mail with a letter of invitation explaining the objectives and procedure of the study to a total of 29,417 eligible subjects between April 1996 and May 1998. The subjects were asked to complete the questionnaire at the worksite or at home and return it in a thick envelope with their written consent to participate to occupational health division. Information on their IDs was also collected for a future linkage with medical records. A total of 25,104 questionnaires were returned. The average response rate was 85%, ranging from 73% to 100% at most study sites, with an exception (43%) at one site. We excluded 3,026 responses from one study site collected during the health checkups between June 1997 and November 1997 because the questionnaire that was distributed during that period lacked a part of the JCQ scales due to an editorial mistake. In addition, 2,421 respondents were excluded because of at least one missing response for variables relevant to the study. Furthermore, a small number of respondents (n=135) who reported their occupation as farming or in a miscellaneous category were excluded from the following analyses. The data from 19,522 respondents (16,444 men and 3,078 women) were analyzed. Classification of occupationThe ISCO88 classifies and ranks occupations according to the levels of skills required and education needed to perform a particular occupation.11 Legislators and managers were ranked the highest, followed by professionals, technicians, clerks, service and sales workers, craft and related trade workers, and machine operators and assemblers. Laborers were ranked the lowest. Respondents were asked to briefly describe their job titles and their most important roles at work, as well as to select their occupation from a multiple-choice question. Based on the descriptions, a four-digit occupational code was entered from the ISCO88 by trained raters under the supervision of researchers (TH, NK). In this study, the first digit of the ISCO88 occupation code (except for military and agricultural/fishery occupations) was used to determine eight occupation categories: managers, professionals, technicians, clerks, service and sales workers, craft and related trade workers, machine operators and assemblers, and laborers. A thorough review of the respondent's description and coding revealed a problem with this procedure: a number of respondents who rated themselves as having managerial occupations (a section chief or a higher position) in the multiple-choice question regarding occupation did not mention their positions as managers in the description section. Thus, their occu...

Positive affect items with positive wording cannot be used to assess depressive disorders in the Japanese population adequately, but this can be done with the corresponding negatively revised items.

The objective of the present study is to investigate the association between various occupational stressors and heavy drinking among male Japanese workers in different age groups. Using the Generic Job Stress Questionnaire, 13 occupational stressors and 2 workplace support indicators were assessed. The questionnaire survey was conducted of 25,104 workers, and the present study analyzed the data from 17,501 male workers. Heavy drinking was defined as weekly alcohol consumption of >275 g, and a total of 1,131 men (6.5%) were classified as heavy drinkers. After adjusting for shift work, occupational class, marital status and smoking, heavy drinking was related to "support from supervisor" for the 18-29 and 50-72 yr-old groups. For the 30-39 yr-old group, heavy drinking was related to "intragroup conflict", "job control" and "cognitive demands." For the 40-49 yr-old group, heavy drinking was related to "physical environment", "quantitative workload" and "underutilization of abilities." The present study clarified that certain occupational stressors relate to heavy drinking, and that this association varies among different age groups.

Job control and role ambiguity may be important predictors of long-term sick leave due to depressive disorders among male employees, independent of depressive symptoms and neuroticism.