The present study formed a part of the longitudinal Kuopio Depression Study (KUDEP), which was conducted in the district of Kuopio in the central-eastern part of Finland [19
]. A sample of the general population (N = 3004) living in Kuopio, aged 25-64 years, was randomly selected from the National Population Register. The study protocol was approved by the Research Ethics Committee of Kuopio University Hospital and the University of Kuopio. All subjects provided written informed consent before entering the study.
A baseline study questionnaire was mailed in 1998 (T1) with a response rate of 68% (n = 2050). In addition, two subsequent follow-up health questionnaires were mailed in 1999 (T2) (n = 1722) and 2001 (T3) (n = 1593). A total of 1347 study subjects responded to all three data collections. In 2005 (i.e. seven years after the baseline), a sub-sample of the subjects was invited for clinical evaluation.
The inclusion criteria for the 2005 sub-sample were based on the presence or absence of self-reported adverse mental symptoms. The following cut-offs were used: BDI-21 score > 9, TAS-20 score > 57 or LS-4 score > 11. Half of the subjects fulfilled at least one of these criteria in all three data collection years (n = 209). A second group with a similar size and the same age and gender distribution was formed from those subjects who were asymptomatic with respect to any of these adverse symptoms during the follow-up (n = 218). The participation rate for the sub-sample was 78% (total n = 333). Out of these, three cases with inadequate data on life satisfaction were excluded. Thus, the final sub-sample comprised 330 subjects. The mean age was 56.4 years (SD = 9.6) for men (n = 142; 43%) and 55.4 years (SD = 9.5) for women (n = 188; 57%) (p = ns).
Life satisfaction was assessed with a self-reported 4-item life satisfaction scale (LS
, range 4-20, higher scores indicating lower life satisfaction), which was originally modified from quality of life studies [20
]. Study subjects were asked to rate their general interest and happiness in life, ease of living and experiences of loneliness with the following scores, respectively: 1 = very interesting/happy/easy/not at all lonely; 2 = fairly interesting/happy/easy; 3 = cannot say/missing data; 4 = fairly boring/unhappy/hard/lonely; 5 = very boring/unhappy/hard/lonely.
The life dissatisfaction burden, an indicator of long-term life satisfaction, was based on the sum of LS scores in the first three follow-up assessments (T1 to T3). The study subjects were divided into tertiles according to this burden: the long-term life satisfaction group1 [range 12-21; n = 116 (35.2%)], intermediate group2 [range 22-33; n = 107 (32.4%)] and long-term dissatisfaction group3 [range 34-60; n = 107 (32.4%)].
The health questionnaires included questions concerning the socio-demographic background. Age, gender, marital status (cohabiting i.e. married or living with a partner vs. non-cohabiting i.e. unmarried or living without a partner), subjective economic status (good/fairly good vs. fairly poor/poor), subjective working ability (good vs. reduced/unable to work), current smoking (yes/no) and alcohol consumption (2-3 times per week/more vs. once a week/less) were recorded.
Depressive symptoms were rated using the self-administered 21-item Beck Depression Inventory (BDI-21
, range 0-63) [23
] and the 17-item Hamilton Depression Rating Scale (HDRS
, range 0-52) [24
] assessed by a trained nurse. The alexithymic symptoms were screened using the 20-item Toronto Alexithymia Scale (TAS-20
; range 20-100) [25
]. Hopelessness was assessed by the 20-item Beck Hopelessness Scale (HS-20
; range: 0-20) [28
]. The 12-item self-reported General Health Questionnaire (GHQ-12
; range 0-36) [29
] was used to assess mental distress. The 28-item self-reported Dissociative Experiences Scale (DES-28
; range 0-100) [30
] was used to assess psychological dissociation. In all of these psychometric scales, higher scores indicated higher psychopathology.
In 1999, self-reported data on previous physician-diagnosed major depressive disorder (MDD) was obtained from the health questionnaire. In 2005, the diagnoses of major depressive disorder were assessed by means of the Structured Clinical Interview for DSM-IV (SCID-I) [31
Data analysis was carried out with SPSS (version 17.0). The differences between the study groups were examined with the Pearson chi-squared test for categorical variables, and analysis of variance (ANOVA) for continuous variables. In the case of variables not following a normal distribution, the non-parametric Kruskal-Wallis test was used.
Logistic regression (method:enter) was used to prospectively examine how the life satisfaction burden in 1998-2001 was related to a subsequent diagnosis of major depression disorder in 2005, and how MDD in 2005 was related to the previous long-term life satisfaction burden. In the logistic models, the LS burden was mostly treated as a continuous score, but as an outcome variable the dissatisfied group was compared with the satisfied group (LS 34-60 vs. 12-21). The final logistic regression models were adjusted for possible confounders (factors significantly associated with LS burden) and they were carried out with and without including major depressive disorder (MDD) diagnosed by a physician in 1999 in order to investigate the associations of new cases with MDD.