Name: SMOKE JEM
The JEM is based on representative Danish surveys with almost 300,000 participants in the period from 1973 to 2013, where information were collected in consistent ways and with information on occupation at the time of survey participation. The lifestyle JEM is computed in random intercept multilevel models using BLUP estimators by DISCO-88 code.
Contact information:Esben Meulengracht Flachs, Dept. of Occupational and Environmental Medicine, Copenhagen University Hospital, Bispebjerg, Bispebjerg Bakke 23, 2200 Copenhagen NV, Denmark. Mail: esben.meulengracht.flachs@regionh.dk
Reference: Petersen SB, Flachs EM, Prescott E, et al. Job-exposure matrices addressing lifestyle to be applied in register-based occupational health studies. Occup Environ Med 2018; 75: 890-7.
Access: To obtain access to data follow the application procedure on this site. You will find an overview of the access posibilies to the JEM here Hent dokument.
Exposure: Smoking (proportion) and smoking (volume)
Coverage: All DISCO-88 (ISCO-88) codes covered with age-period-sex specific estimates.
Job title and industry classification system: DISCO-88.
Calendar period: 6 time-periods: 1981-1990, 1991-1995, 1996-2000, 2001-2005, 2006-2010, >2010
Measures of exposure: Which measures are used to describe exposure?
- SMOKE JEM (P): probability of active smoking, where previous smokers are defined as non-smokers.
- SMOKE JEM (V): gram of tobacco smoked per day.
Uncertainty in exposure measures: At least 10 observations are used for the estimation. If 10 observations not have been available then the DISCO-code from a lower level is used in the JEM. Those DISCO-codes can be identified as they are listed with an extra variable for FAGKODE numbered with a lower level.
Demographic dimensions: Men and women from 15 years of age (average= ). Data underlying the JEM is mostly from the areas in/around Copenhagen or Aarhus in Denmark in the first time-periods. I the two recent time-periods data should be representative for the whole country.
