Study population
Data from eight ongoing population cohort studies in Sweden, with baseline investigations performed between 1973 and 2003, were used for this study. The study population has previously been described in detail [23, 24]. Briefly, random samples based on predefined specifications concerning age, sex, and area of residence, were drawn from the national population register. The Men Born in 1913 subsample consisted of a random third of the male population aged 60 in the city of Gothenburg, Sweden, in 1973, and the Men Born in 1923 subsample consisted of a random tenth of the male population aged 50 in Gothenburg Table 1. Survivors in the samples were invited to re-examinations in 1980, 1988 and 1993.
The Men Born in 1943 subsample consisted of a random third of 50-year-old men living in Gothenburg in 1993, re-examined in 2003, and the Women and Men Born in 1953 subsamples consisted of a random third of 50-year old women and men living in Gothenburg in 2003. The Eskil subsample consisted of a random sample of 625 men aged 30–54 and living in the city of Eskilstuna, Sweden in 1986. The Uppsala Public Health Cohort was based on random samples of 1000 women and men 25 years or older from each of the six municipalities of Uppsala County in 1993. The Beda II subsample was based on a re-examination in 1997 of a random sample drawn in 1979 of 1746 women born 1915–1941 and living in Gothenburg. The Uppsala-Örebro Women Study sample was based on random samples of 600 women aged 35–64 from each of the seven counties in the Uppsala-Örebro Health Care Region, Sweden.
All samples were by definition representative of their underlying general populations. No exclusions were made. The combined samples consisted of 20,160 subjects of whom 3,590 were part of more than one subsample. Overall, 14,470 (71.8%) observations were obtained, based on 12,000 unique individuals. Of these 10,451 (6,808 women and 3,644 men) participated once, 964 men twice, 254 men three times, and 330 men on four occasions. Additional information on the study population is given in Table 1.
Ethical consideration
Informed consent was obtained from all participants, verbal in the early studies, and written later on, as required first by the Research Ethics Committees at Gothenburg and Uppsala Universities, and later by the National Research Ethics Board. The Committees and the Board approved the study on several occasions during the data collection process.
Data collection
Outcome data were obtained from official registers. Other data used in this report was obtained from baseline or follow-up examinations, in some of the studies by questionnaire in connection with medical examinations performed, in others by postal questionnaires.
Educational level was classified on a four-point scale ranging from ‘compulsory education only’ (=1), to ‘college or university level education’ (=4). Employment status was measured on a four-point nominal scale as ‘gainfully employed’ (including students), ‘unemployed’, ‘on sick leave or disability pension’, or ‘old age retirement’. Marital status was classified as married/ cohabiting or not (the latter including response alternatives never married, divorced, and widowed).
Self-rated health (SRH) was measured with the Well-being subscale of The Gothenburg Quality of Life instrument (GQL) [25]. Respondents were asked to rate their health on a seven-point Likert scale with response alternatives ranging from ‘very bad’ (=1) to ‘excellent, could not be better’ (=7), and with no verbal labels of the intervening steps. The seven-point scale was used in the analyses. For illustration purposes in the Figures the seven-point scale was converted into a three point scale (1–3 (11%), 4–5 (35%), and 6–7 (54%)). Symptom reporting was assessed based on the Complaint Score subscale of the GQL, in which subjects are asked ‘Have you been troubled by any of the following symptoms during the past three months?’, followed by a list of 30 symptoms with response alternatives ‘yes’ (=1) or ‘no’ (=0) given for each symptom. The Complaint Score was obtained as the sum across the 30 symptoms.
Leisure time physical activity was reported on a four-point ordinal scale with response alternatives ‘sedentary’, ‘moderately active’, ‘active’, or ‘vigorously active’ [26]. Smoking habits were classified as ‘current smoker’ or ‘non-smoker’ (including never smoked and ex-smoker). In addition, in some of the cohort studies a five-point smoking variable was available, where smoking habits were classified as ‘never smoked’ (=1), ‘ex-smoker’ (=2), ‘currently smoking 1–14 grams of tobacco per day’ (=3), ‘smoking 15–24 grams per day’ (=4), or ‘smoking 25 grams or more per day’, one cigarette equalling 1 gram, one cheroot 2 grams, one cigar 5 grams, and pipe tobacco 50 grams divided by the number of days the pack lasted [26].
The Swedish Social Insurance Agency administers all sick leave and disability pension benefits, and their database is a complete account of official sick leave compensation and disability benefits granted. Information on all compensated days of sick leave for each individual in the study populations from 1 January 1986 until 31 December 2002 was obtained from the Agency.
The data included the first and last day of each sickness spell, the type of sick leave benefit (compensation for sickness, work injury, or rehabilitation), and extent of sick leave (25%, 50%, 75% or 100%).
Information on whether the subjects had been granted a disability pension at any time from 1971 until 2001 was obtained from the Agency. The data included decision date, diagnoses, extent (25%, 50%, 67%, 75% or 100%) and type (temporary or permanent) of disability pension. The incidence rate of new disability pension was calculated among those who did not have a disability pension at baseline.
Data on all hospital admissions from 1971 until December 31, 2002 was obtained from the National Hospital Discharge Register. The data obtained included date of admission, date of discharge and diagnoses. In this report only the main diagnosis was used. Data on cause-specific mortality from 1971 until December 31, 2002 was obtained from the National Causes of Death Register. The data used here was date of death, and underlying cause of death. The disability pension diagnoses, discharge diagnoses, and causes of death were classified according to the International Statistical Classification of Diseases and Related Health Problems (ICD) versions 8–10.
Statistical consideration
Data was analyzed with the Statistical Analyses System software (SAS) version 9.2 [27]. Data concerning age, sex, examination year, and outcome data were complete, except for one individual where age was missing. Not all variables were measured in all subpopulations. Complaint score was not measured in the Uppsala-Örebro Women Study, and leisure time physical activity was not measured in the Eskil Study. The number of available observations for each variable is shown in Table 2. The overall proportion of missing data in subpopulations where the variables were measured was less than 2%. Missing data were not replaced. Simple differences between groups were assessed with Student’s t-test or the chi-square test.
Multiple linear regression, performed with the SAS procedure ‘GLM’ providing regression coefficients, 95% confidence intervals, and Wald’s chi-square (a measure of variable impact, and p-values), was used in the analyses with the dependent variable as the outcome (number of sick leave days during follow up), and SRH and the covariates age, examination year, marital status, smoking habits, physical activity during leisure time, educational level, being unemployed, and complaint score as independent variables, with backward elimination of non-significant variables. Moreover, possible nonlinear relationships were tested with squared and cubic terms, and so were possible interaction terms but none was found.
Proportional hazards regression (Cox’s analysis) was used in the analyses of the effects of SRH at baseline on the outcome variables survival, admission to hospital, and being granted a disability pension, with the outcome entered as the dependent variable and with SRH and the same covariates as mentioned above, and being on sick leave or disability pension (regarding the outcomes hospital admission and mortality) as independent variables using the SAS procedure ‘Phreg’ providing hazards ratios (HR), 95% confidence intervals (CI), Wald’s chi-square and p-values).
The choice of independent variables potentially associated with outcome was based on variables at hand in most of the cohorts (a substantial number), and on the literature. In addition, some were chosen based on intuition. The relationship between the outcome variables and the candidate independent variables was tested in bivariate analyses stratified by sex. All variables related to an outcome in women or men were entered in the final analytical models.
Since the proportional hazards regression is dependent of proportionality between the hazard function of those exposed and those not exposed, the hazard functions across the follow-up period were calculated separately for men and women with the SAS procedure ‘Life-test’ providing tabulated and plotted hazards levels for exposed and unexposed groups. The analyses of hospital admission and disability pension were adjusted for non-exposure by censoring subjects at time of death. All tests were two-tailed. Significance levels were set at P<0.05.