Örebro University
School of Medical Sciences Degree project, 15 credits June 2020
Inflammation in young Swedish men and risk of
adult-onset depression defined by prescription of
antidepressant medications
Author: Simon Ghanem Supervisors: Scott Montgomery1, 2, 3, Ayako Hiyoshi1, Katja Fall1,2
1 Clinical Epidemiology and Biostatistics, Faculty of Medicine and Health, Örebro University, SE-701 82 Örebro, Sweden
2 Clinical Epidemiology Unit, Department of Medicine, Karolinska Institutet, SE-171 76 Stockholm, Sweden 3 Department of Epidemiology and Public Health, University College London, 1-19 Torrington Place, WC1E 6BT London, United Kingdom
Abstract
Background: Depression is a major disability that has been studied extensively in regard to systemic inflammation. Cross-sectional studies have shown that systemic inflammatory parameters are raised in individuals with depression, but few studies have explored the longitudinal direction of associations. The aim of this study is to investigate whether
heightened ESR-levels during adolescence increases the risk of subsequent depression during adulthood measured by antidepressant medication.
Methods: This register-based cohort study included 196,667 Swedish men who were born in 1952 to 1956, attended mandatory military conscription assessments in late adolescence, and were followed up to 2009 through linkages of various national registers. The association between erythrocyte sedimentation rate, measured at conscription examination, and antidepressant treatment in middle age was examined using logistic regression with
adjustment for confounding by BMI, stress resilience, socioeconomic index and household crowding.
Results: Erythrocyte sedimentation rate was not positively associated with an increased subsequent risk of antidepressant treatment. The characteristics of >18,5 BMI, low stress resilience, lower household crowding and the socioeconomic index belonging to ‘office worker’, were of higher risk for antidepressant treatment.
Conclusion: This study shows that the systemic inflammatory marker erythrocyte
sedimentation rate is not positively associated with subsequent development of depression defined by antidepressant treatment.
INTRODUCTION 1
METHODS 2
Study population 2
Main exposure 2
Outcome 2
Potential confounding factors 3
Statistical analysis 3 RESULTS 4 Table 1 6 Table 2 6 DISCUSSION 7 CONCLUSION 8 REFERENCES 9
Introduction
Depression is a major disability responsible for most years lost to disease worldwide [1] with a prevalence of 6,9% in the EU [2] and is linked with increased mortality [3–5] and large economic impact [6,7]. The association between systemic inflammation and depression has been studied extensively using cross-sectional studies, but few have studied inflammation and its longitudinal role in subsequent development of depression. To map the interplay between depression and inflammation can be key in understanding the course of this disease and future opportunity of treatment.
Cross-sectional studies that have been summarized in meta-analyses point towards increased levels of inflammatory markers, such as C-reactive protein (CRP), IL-6, TNF- and sIL-2R [8–10], in individuals with depression. Oppositely, some studies have shown that not all depressed patients present with raised inflammation [11,12]. Moreover, some findings indicate only a particular subset of depressed patients develop raised inflammatory markers [13], such as individuals who has experienced child adversity [14], cumulative episodes of depression [15] and non-atypical depression [16]. There is also evidence that elevated CRP could be explained by confounding factors [17,18]. Despite some controversy, the majority of studies indeed do point towards an association, but whether inflammation causes depression or if depression causes inflammation is less clear.
Few studies have examined the direction of associations between inflammation, using primarily CRP and IL-6 as inflammatory markers, and depression and these studies have presented conflicting results [19–25]. Erythrocyte sedimentation rate (ESR), an indirect measure of acute-phase proteins that is affected by numerous factors [26], has shown to be higher in individuals with depression; not only in comparison to control-groups [27], but also in correlation to severity [28]. To the best of our knowledge only one study has examined the longitudinal link between ESR and subsequent risk of depression and concluded that ESR is not associated with depression in 746 veterans with a follow-up period of four years [29]. However, these studies using prospective data to explore the direction of associations are apart from being few in numbers, relatively restricted with regard to study size, follow-up period and rely on a questionnaire-based identification of depression, all of which might be part of reasons of inconsistent findings. It is thus evident that this field still contains uncertainty, mainly in the direction of associations, and needs further studying to advance the understanding of depression and its pathology.
2 The aim of this study is to investigate whether young Swedish men who had an increased ESR-baseline during military conscription had an increased risk of subsequent depression defined by a prescription of antidepressant medication equal to or more than 180 days within 365 days in middle age using prospective data.
Methods
Study population
The study population is a cohort of men born between 1952 to 1956 who attended mandatory military conscription in Sweden when the majority of them were aged 18. Fewer than 3% of men were exempted from conscription due to severe disease or disability [30]. Of the 284,238 identified men aged 17 to 20 at the time of conscription, we excluded individuals with invalid vital status (n=2623) and diagnosis of psychiatric disease at the time of conscription (n=34,602). We further excluded those who had emigrated (n=15034) or died (n=14878) before the start of observation, 1 January 2006; finally resulting in 196,667 individuals who were analyzed for this study. Conscription registers were provided by the Swedish National Archives
(https://riksarkivet.se/using-the-archives).
Approval of ethics for this study has been obtained from Uppsala Regional Ethics Committee (Dnr 2014/324).
Main exposure
At conscription examination trained personnel sampled blood for analysis to allow for measurement of ESR and erythrocyte volume fraction (EVF). ESR was divided into four equal sized groups, from low to high levels of inflammation (Table 1). ESR was always adjusted for EVF to account for individual variations in blood composition.
Outcome
To identify individuals who developed depressive disorders later in middle age, we linked the cohort with the Swedish Prescription Register, issued by the National Board of Health and Welfare (https://www.socialstyrelsen.se/en/), which was launched in July 2005 and contains patient data and all prescribed drugs for the whole population [31]. Using the defined daily dose (DDD)
described by the World Health Organization, which approximates the assumed average daily intake for a drug, we defined the incidence of depressive order by approximately continuous treatment of antidepressant medication more than or equal to 180 days within 365 days.
Potential confounding factors
ESR is a non-specific marker of inflammation that can be raised due to numerous reasons [26]. For this reason, we adjusted for several possible confounders. At conscription examination participants were assessed for their physical, psychiatric and socioeconomic factors. Early stage of the examination a questionnaire with questions on aspects of familial, medical, social behavior and personality characteristics was filled by potential conscripts. A psychologist evaluated stress resilience using a semi-structured interview complemented by the self-filled questionnaire summarized into a nine-level normally distributed score, which was collapsed into three groups of high, intermediate and low in the present study. Recording of medical and psychiatric disorders based on ICD-8 were assessed by a physician. Trained personnel measured height and weight. Socioeconomic variables were defined based on socioeconomic index (SEI) and household crowding, derived from census in 1960, which were supplied by the government agency Statistics Sweden (https://www.scb.se/en/services/) that holds all population-based registers in Sweden. SEI was measured by the household head’s occupation and categorized into six groups. Household crowding was calculated by dividing the number of household members by the number of habitable rooms and grouped into the following: less than one, one to two or more than two persons per room.
Statistical analysis
Cross-tabulation was used to describe the study population and its characteristics, those who were depressed, and inflammatory levels. Logistic regression is a type of statistical method that models the odds of a binary outcome for one or more explanatory factors. It compares odds of each category to a reference category yielding an odds ratio (OR). OR is the relative odds of the outcome event to occur in relation to the exposure. OR<1 indicate the exposure is associated with lower odds of outcome, OR>1 indicate the exposure is associated with higher odds of outcome, and OR=0 suggests the exposure does not affect the outcome. Confidence interval (CI) shows the interval within which the true value of the population would lie, in this study a 95% CI was used.
4 Unadjusted logistic regression, considering one explanatory factor in relation to the outcome, was performed to examine crude associations between the exposure and potential confounding factors with the outcome, producing the unadjusted ORs. Multivariable logistic regression includes potential confounding factors, together with the exposure, in the model. Therefore, ORs produced by this analysis were adjusted for potential confounding by BMI, stress resilience, SEI and household crowding. The statistical analysis was conducted with IBM SPSS Statistics 26.
Results
Table 1 presents the study population’s characteristics, those who were depressed, and low (1) to high (4) ESR-levels during adolescence. Table 2 presents the unadjusted and adjusted ORs for the outcome with 95% CIs. The ORs show a modest inverse association between inflammation and depression in both the unadjusted (0.98 [0.94–1.03], 0.95 [0.90–1.00], 0.96 [0.90–1.01]) and the adjusted (0.99 [0.94–1.04], 0.95, [0.90–1.00], 0.95 [0.90–1.01]) model which accounts for potential confounding by BMI, stress resilience, SEI and household crowding. Only ESR-level 3 in the unadjusted model was statistically significant (P-value = 0.04). Further, in the adjusted model, significant associations were found in some categories of BMI, stress resilience, parental socioeconomic index and parental household crowding. Underweight individuals were more likely to be prescribed antidepressants than normal weighing individuals (1.09 [1.03–1.15]). Those with lower and moderate stress resilience were of higher risk for receiving antidepressants compared to high stress resilience individuals, with ORs 1.62 [1.53–1.71] and 1.19 [1.14–1.25], respectively. In the SEI category, compared to the ‘business worker’ group, participants belonging to the ‘office worker’ category were more likely (1.10 [1.04–1.17]) to be associated with antidepressant medication and participants belonging to the ‘farm owner’ category were less likely (1.14 [0.79–0.93]). Individuals from households with 1–2 and >2 persons per habitable room were less likely to be associated with the outcome (1.09 [0.85–0.98] respectively 1.08 [0.85–0.99]). Very similar ORs were present in both unadjusted and adjusted models when the potential confounding factors were accounted for. Lastly, almost identical statistical significance appeared in both models, except for an additional significant value in the category of ESR-level 3, unadjusted model
Additionally, a post hoc analysis (not presented in the tables) by each group of SEI was conducted using multivariable logistic regression. ORs for ESR in participants belonging to the ‘business owner’, ‘office owner and ‘others’ showed no clear pattern in ESR. Participants
belonging to the ‘farm owner’ and ‘agricultural worker’ showed a consecutive increase in likelihood of antidepressant treatment with increasing ESR levels: 0.913, 0.946, 1.027 for ‘farm owner’, and 0.862, 0.919, 1.003 for ‘agricultural worker’, albeit without significance.
6 Table 1
Characteristics of the study population, depression and ESR-levels.
Total cohort Depression ESR 1 (lowest) ESR 2 ESR 3 ESR 4 (highest) N = 196667 N = 13418 N = 46000 N= 56629 N = 53800 N = 40238 Freq. (%) Freq. (%) Freq. (%) Freq. (%) Freq. (%) Freq. (%)
ESR (adjusted for EVF)
1 (lowest) 46000 (23.4) 3293 (24.5) 2 56629 (28.8) 3920 (29.2) 3 53800 (28.8) 3553 (26.5) 4 (highest) 40238 (20.5) 2652 (19.8) BMI Underweight 21987 (11.2) 1673 (12.5) 4604 (10.0) 6189 (10.9) 6056 (11.3) 5138 (12.8) Normal 160181 (81.4) 10779 (80.3) 37812 (82.2) 46401 (81.9) 43953 (81.70) 32015 (79.6) Overweight 12654 (6.4) 824 (6.1) 3136 (6.80) 3600 (6.4) 3303 (6.1) 2615 (6.5) Obese 1845 (0.9) 142 (1.1) 448 (1.0) 439 (0.8) 488 (0.90) 470 (1.2) Stress resilience Low (1–3) 27931 (14.2) 2505 (18.7) 6623 (14.4) 7867 (13.9) 7495 (13.9) 5946 (14.8) Moderate (4–6) 116768 (59.4) 7904 (59.9) 27654 (60.1) 33595 (59.3) 31586 (58.7) 23933 (59.5) High (7–9) 51968 (26.4) 3009 (22.4) 11723 (25.5) 15167 (26.8) 14719 (27.4) 10359 (25.7) Parental socioeconomic index
Business owner 21240 (10.8) 1435 (10.7) 5128 (11.1) 6112 (10.8) 5725 (10.6) 4275 (10.6) Office worker 54785 (27.9) 4000 (29.8) 13189 (28.7) 16174 (28.6) 14991 (27.9) 10431 (25.9) Farm owner 20831 (10.6) 1228 (9.2) 4364 (9.5) 5887 (10.4) 5927 (11.0) 4653 (11.6) Manual worker 80811 (41.1) 5460 (40.7) 18836 (40.9) 23002 (40.6) 22010 (40.9) 16963 (42.2) Agricultural worker 7728 (3.9) 495 (3.7) 1820 (4.0) 2140 (3.8) 2126 (4.0) 1642 (4.1) Others 11272 (5.7) 800 (6.0) 2663 (5.8) 3314 (5.9) 3021 (5.6) 2274 (5.7) Parental household crowding
<1 12515 (6.4) 907 (6.8) 3091 (6.7) 3642 (6.4) 3383 (6.3) 2399 (6.0) 1-2 104149 (6.4) 7036 (52.4) 24644 (53.6) 30028 (53.0) 28501 (53.0) 20976 (52.1) >2 80003 (40.7) 5475 (40.8) 18265 (39.7) 22959 (40.5) 21916 (40.7) 16863 (41.9)
Table 2
Adjusted and unadjusted ORs with fitting P-value and 95% CI.
Unadjusted P-value Adjusted P-value
OR 95% CI OR 95% CI
ESR (adjusted for EVF)
1 (lowest) ref. ref.
2 0.98 [0.94–1.03] 0.47 0.99 [0.94–1.04] 0.55*
3 0.95 [0.90–1.00] 0.04 0.95 [0.90–1.00] 0.06*
4 (highest) 0.96 [0.90–1.01] 0.12 0.95 [0.90–1.01] 0.12*
BMI
Underweight 1.14 [1.08–1.20] < 0.001 1.09 [1.03–1.15] < 0.001
Normal ref. ref.
Overweight 0.97 [0.90–1.04] 0.35 0.95 [0.89–1.03] 0.20
Obese 1.11 [0.97–1.37] 0.10 1.09 [0.92–1.30] 0.32
Stress resilience
Low (1–3) 1.60 [1.52–1.69] < 0.001 1.62 [1.53–1.71] < 0.001
Moderate (4–6) 1.18 [1.13–1.13] < 0.001 1.19 [1.14–1.25] < 0.001
High (7–9) ref. ref.
Parental socioeconomic index
Business owner ref. ref.
Office worker 1.09 [1.02–1.16] 0.01 1.10 [1.04–1.17] < 0.001
Farm owner 0.87 [0.80–0.94] < 0.001 0.86 [0.79–0.93] < 0.001
Manual worker 1.00 [0.94–1.06] 1.00 0.98 [0.92–1.04] 0.55
Agricultural worker 0.95 [0.85–1.05] 0.29 0.91 [0.82–1.01] 0.09
Others 1.05 [0.96–1.15] 0.25 1.03 [0.94–1.13] 0.55
Parental household crowding
<1 ref. ref.
1-2 0.93 [0.86–1.00] 0.04 0.91 [0.85–0.98] 0.02
>2 0.94 [0.87–1.01] 0.1 0.92 [0.85–0.99] 0.03
Discussion
This large cohort of 196,667 Swedish men who attended mandatory conscription assessments between 1969 and 1976, aged 17 to 20, showed a modest decrease in subsequent depression among those with higher levels of inflammation in late adolescence. In contrast, one study [29] examining 746 veterans over a period of four years, concluded that ESR was not linked with subsequent risk of depression.
This study benefits from its very large participant number being close to 200,000, a reasonably fair sample of Swedish men at the time considering conscription was mandatory and fewer than 3% were excluded from being enrolled and a follow-up of at least 32 years. In addition, we adjusted for potential confounding factors that are associated with elevated ESR and used a national prescription registry with a record of every dispensed drug.
The SEI category ‘farm owner’ and ‘agricultural worker’ showed an unexpected association by having the lowest likelihood for antidepressant treatment. Observe that the group ‘agricultural worker’ did not yield statistically significant values in the analysis, the group is however discussed in conjunction with the ‘farm worker’ group due to their similar patterns in regard to depression, inflammatory levels and homogeneity that is discussed further down. Farmers and agricultural workers are in general particularly prone to stressors [32], high levels of depression [33] and high rates of suicide when compared to other occupational groups [34]. It therefore seems unlikely that these findings translate over to a lower rate of depression among the participants belonging to this group in comparison to the other groups. A noteworthy detail concerning SEI is that the measure reflects the household head’s occupation, and not directly the participant’s character. However, parental education is indeed an important factor in child achievement and wellbeing [35]. The post hoc analysis, an effort to adjust for this discrepancy in the SEI, did not yield any decisive results leaving us with a possible case of residual confounding.
Lower stress resilience was associated with higher risk of subsequent antidepressant treatment, being consistent with earlier studies [36,37]. Individuals of <18,5 BMI were more likely to be on antidepressant medication compared to individuals within a normal BMI. Such an association, between underweight individuals and depression, has been shown in some studies [38,39]. Higher crowding was associated with lower likelihood of antidepressant treatment when compared to lower crowding. It is however questionable to interpret this finding as if there was a lower occurrence of depression among participants of higher crowding since higher crowding has been associated with adverse effects on mental health [40]. Lastly, the
8 group ‘office worker’ was the most likely SEI associated with antidepressant treatment. It seems questionable that this particular group would have the greatest occurrence of depression, given the associations between parental education and child outcome [35].
Limits to this study include the facts that this cohort contains only men, measurement of ESR was only during adolescence and registration of antidepressant medication occurred during the participants 50’s why these findings may not be applicable to women, youngsters or older groups. The recording of prescriptions was only open for a brief period of time, 2006 to 2009. Furthermore, measuring pharmaceutical dispensing to estimate chronic disease, such as depression, in a population is a common method used in epidemiological studies [41–43], but this way of estimation does comprise some weaknesses [44]. Indications for being prescribed antidepressants include a variety of reasons other than depression [45,46]. To avoid inclusion of individuals who were potentially treated with antidepressant medication for other reasons than depression we included only participants on antidepressant treatment equal to or more than 180 days within 365 days. Six months is the minimum recommended duration of treatment for depression after clinical recovery [47] and thus more likely indicating an actual case of depression rather than a false positive.
Limitations of this study more problematic to account for are cultural and social views on antidepressant treatment. Older adults have shown to be more reluctant to accept antidepressant medication [48,49], and this effect is further accentuated in farmers and agricultural workers [34,50]. We hypothesize that his may have been a factor in the low prescription of antidepressants to farmers and agricultural workers and perhaps a component in the unexpected modest inverse association found in the analysis.
Conclusion
A growing body of evidence suggests that systemic inflammatory markers are raised in individuals afflicted with depression and a few longitudinal studies suggest that some of these markers, such as CRP and IL-6, are associated with subsequent development of depression. This study shows that the inflammatory marker ESR is not positively associated with an increase in depression defined by a measure of prescription of antidepressants.
References
1. Smith K. Mental health: a world of depression. Nature 2014; 515:181.
2. Wittchen HU, Jacobi F, Rehm J, Gustavsson A, Svensson M, Jönsson B, m.fl. The size and burden of mental disorders and other disorders of the brain in Europe 2010. Eur
Neuropsychopharmacol J Eur Coll Neuropsychopharmacol 2011; 21:655–79.
3. Wulsin LR, Vaillant GE, Wells VE. A systematic review of the mortality of depression. Psychosom Med 1999; 61:6–17.
4. Saz P, Dewey ME. Depression, depressive symptoms and mortality in persons aged 65 and over living in the community: a systematic review of the literature. Int J Geriatr
Psychiatry 2001; 16:622–30.
5. Cuijpers P, Smit F. Excess mortality in depression: a meta-analysis of community studies. J Affect Disord 2002; 72:227–36.
6. Sobocki P, Jönsson B, Angst J, Rehnberg C. Cost of depression in Europe. J Ment Health Policy Econ 2006; 9:87–98.
7. Greenberg PE, Fournier A-A, Sisitsky T, Pike CT, Kessler RC. The economic burden of adults with major depressive disorder in the United States (2005 and 2010). J Clin Psychiatry 2015; 76:155–62.
8. Dowlati Y, Herrmann N, Swardfager W, Liu H, Sham L, Reim EK, m.fl. A meta-analysis of cytokines in major depression. Biol Psychiatry 2010; 67:446–57.
9. Liu Y, Ho RC-M, Mak A. Interleukin (IL)-6, tumour necrosis factor alpha (TNF-α) and soluble interleukin-2 receptors (sIL-2R) are elevated in patients with major depressive disorder: a meta-analysis and meta-regression. J Affect Disord 2012; 139:230–9.
10. Howren MB, Lamkin DM, Suls J. Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis. Psychosom Med 2009; 71:171–86.
11. Whooley MA, Caska CM, Hendrickson BE, Rourke MA, Ho J, Ali S. Depression and inflammation in patients with coronary heart disease: findings from the Heart and Soul Study. Biol Psychiatry 2007; 62:314–20.
12. Glassman AH, Miller GE. Where there is depression, there is inflammation... sometimes! Biol Psychiatry 2007; 62:280–1.
13. Lopresti AL, Maker GL, Hood SD, Drummond PD. A review of peripheral biomarkers in major depression: the potential of inflammatory and oxidative stress biomarkers. Prog
Neuropsychopharmacol Biol Psychiatry 2014; 48:102–11.
10 previously exposed to childhood adversity. Biol Psychiatry 2012; 72:34–40.
15. Copeland WE, Shanahan L, Worthman C, Angold A, Costello EJ. Cumulative depression episodes predict later C-reactive protein levels: a prospective analysis. Biol Psychiatry 2012; 71:15–21.
16. Hickman RJ, Khambaty T, Stewart JC. C-reactive protein is elevated in atypical but not nonatypical depression: data from the National Health and Nutrition Examination survey (NHANES) 1999-2004. J Behav Med 2014; 37:621–9.
17. Bjerkeset O, Romild U, Smith GD, Hveem K. The associations of high levels of C-reactive protein with depression and myocardial infarction in 9258 women and men from the HUNT population study. Psychol Med 2011; 41:345–52.
18. Douglas KM, Taylor AJ, O’Malley PG. Relationship between depression and C-reactive protein in a screening population. Psychosom Med 2004; 66:679–83.
19. Osimo EF, Stochl J, Zammit S, Lewis G, Jones PB, Khandaker GM. Longitudinal population subgroups of CRP and risk of depression in the ALSPAC birth cohort. Compr Psychiatry 2020; 96:152143.
20. Gimeno D, Kivimäki M, Brunner EJ, Elovainio M, De Vogli R, Steptoe A, m.fl.
Associations of C-reactive protein and interleukin-6 with cognitive symptoms of depression: 12-year follow-up of the Whitehall II study. Psychol Med 2009; 39:413–23.
21. Khandaker GM, Pearson RM, Zammit S, Lewis G, Jones PB. Association of serum interleukin 6 and C-reactive protein in childhood with depression and psychosis in young adult life: a population-based longitudinal study. JAMA Psychiatry 2014; 71:1121–8. 22. Duivis HE, Kupper N, Penninx BW, Na B, de Jonge P, Whooley MA. Depressive
symptoms and white blood cell count in coronary heart disease patients: prospective findings from the Heart and Soul Study. Psychoneuroendocrinology 2013; 38:479–87.
23. Duivis HE, de Jonge P, Penninx BW, Na BY, Cohen BE, Whooley MA. Depressive symptoms, health behaviors, and subsequent inflammation in patients with coronary heart disease: prospective findings from the heart and soul study. Am J Psychiatry 2011; 168:913– 20.
24. Valkanova V, Ebmeier KP, Allan CL. CRP, IL-6 and depression: a systematic review and meta-analysis of longitudinal studies. J Affect Disord 2013; 150:736–44.
25. Von Känel R, Bellingrath S, Kudielka BM. Association between longitudinal changes in depressive symptoms and plasma fibrinogen levels in school teachers. Psychophysiology 2009; 46:473–80.
Sedimentation Rate and C-reactive Protein Measurements and Their Relevance in Clinical Medicine. WMJ Off Publ State Med Soc Wis 2016; 115:317–21.
27. Chavda N, Kantharia ND, Jaykaran null. Effects of fluoxetine and escitalopram on C-reactive protein in patients of depression. J Pharmacol Pharmacother 2011; 2:11–6. 28. Chang C-C, Tzeng N-S, Kao Y-C, Yeh C-B, Chang H-A. The relationships of current suicidal ideation with inflammatory markers and heart rate variability in unmedicated patients with major depressive disorder. Psychiatry Res 2017; 258:449–56.
29. Eswarappa M, Neylan TC, Whooley MA, Metzler TJ, Cohen BE. Inflammation as a predictor of disease course in posttraumatic stress disorder and depression: A prospective analysis from the Mind Your Heart Study. Brain Behav Immun 2019; 75:220–7.
30. Otto U. Male youths. A sociopsychiatric study of a total annual population of Swedish adolescent boys. Acta Psychiatr Scand Suppl 1976; 264:1–312.
31. Wettermark B, Hammar N, Fored CM, MichaelFored C, Leimanis A, Otterblad Olausson P, m.fl. The new Swedish Prescribed Drug Register--opportunities for
pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiol Drug Saf 2007; 16:726–35.
32. Simkin S, Hawton K, Fagg J, Malmberg A. Stress in farmers: a survey of farmers in England and Wales. Occup Environ Med 1998; 55:729–34.
33. Torske MO, Hilt B, Glasscock D, Lundqvist P, Krokstad S. Anxiety and Depression Symptoms Among Farmers: The HUNT Study, Norway. J Agromedicine 2016; 21:24–33. 34. Gregoire A. The mental health of farmers. Occup Med Oxf Engl 2002; 52:471–6. 35. Chen Q, Kong Y, Gao W, Mo L. Effects of Socioeconomic Status, Parent–Child Relationship, and Learning Motivation on Reading Ability. Front Psychol 2018; 9:1297. 36. Hiyoshi A, Udumyan R, Osika W, Bihagen E, Fall K, Montgomery S. Stress resilience in adolescence and subsequent antidepressant and anxiolytic medication in middle aged men: Swedish cohort study. Soc Sci Med 1982 2015; 134:43–9.
37. Jacobs RH, Reinecke MA, Gollan JK, Kane P. Empirical evidence of cognitive vulnerability for depression among children and adolescents: A cognitive science and developmental perspective. Clin Psychol Rev 2008; 28:759–82.
38. Carpenter KM, Hasin DS, Allison DB, Faith MS. Relationships between obesity and DSM-IV major depressive disorder, suicide ideation, and suicide attempts: results from a general population study. Am J Public Health 2000; 90:251–7.
12 40. Evans GW. The built environment and mental health. J Urban Health Bull N Y Acad Med 2003; 80:536–55.
41. Macchia A, Monte S, Pellegrini F, Romero M, D’Ettorre A, Tavazzi L, m.fl. Depression worsens outcomes in elderly patients with heart failure: an analysis of 48,117 patients in a community setting. Eur J Heart Fail 2008; 10:714–21.
42. Berge LI, Riise T, Fasmer OB, Lund A, Oedegaard KJ, Hundal O. Risk of depression in diabetes is highest for young persons using oral anti-diabetic agents. Diabet Med J Br Diabet Assoc 2012; 29:509–14.
43. Colvin L, Slack-Smith L, Stanley FJ, Bower C. Are women with major depression in pregnancy identifiable in population health data? BMC Pregnancy Childbirth 2013; 13:63. 44. Patten SB, Esposito E, Carter B. Reasons for antidepressant prescriptions in Canada. Pharmacoepidemiol Drug Saf 2007; 16:746–52.
45. Henderson J, Harrison C, Britt H. Indications for antidepressant medication use in Australian general practice patients. Aust N Z J Psychiatry 2010; 44:865.
46. Kwon A, Bungay KM, Pei Y, Rogers WH, Wilson IB, Zhou Q, m.fl. Antidepressant use: concordance between self-report and claims records. Med Care 2003; 41:368–74.
47. Shelton RC. Steps Following Attainment of Remission: Discontinuation of Antidepressant Therapy. Prim Care Companion J Clin Psychiatry 2001; 3:168–74. 48. Jayasekara R, Procter N, Harrison J, Skelton K, Hampel S, Draper R, m.fl. Cognitive behavioural therapy for older adults with depression: a review. J Ment Health Abingdon Engl 2015; 24:168–71.
49. Leaf PJ, Bruce ML, Tischler GL, Holzer CE. The relationship between demographic factors and attitudes toward mental health services. J Community Psychol 1987; 15:275–84. 50. Kallioniemi MK, Simola AJK, Kymäläinen H-R, Vesala HT, Louhelainen JK. Stress among Finnish farm entrepreneurs. Ann Agric Environ Med AAEM 2008; 15:243–9.
