Influence of social support among employees on mental health and work ability—a prospective cohort study in 2013–15

Influence of social support among employees on mental health and work ability—a prospective... Abstract Background The study examined the association of social support with mental health, subjective work ability and psychological consultation. Methods The cohort study included 1886 German employees (40–54 years) with sickness absence exceeding six weeks in 2012. Postal surveys were carried out in 2013 and 2015. Results After adjustment for socio-demographic data, work-related characteristics and personality factors, persons with low social support compared to high social support had poorer mental health (b=−4.96; 95% CI: −7.11 to −2.81) and a lower work ability index (b=−1.10; 95% CI −2.00 to −0.21). Low social support was not associated with increased odds of consulting a psychologist (OR =1.30; 95% CI: 0.86–1.96). Conclusions Low social support is an independent predictor of poorer mental health and lower work ability. This study highlights the importance of identifying people who have limited access to social support. Introduction In Europe, there exists a great demand for effective measures to improve work ability among those threatened by illness. In recent years this applies particularly to mental disorders.1,2 Mental well-being is a key resource for productivity, participation and inclusion. Mental disorders are among the top ten public health challenges in the WHO European Region, affecting about 25% of the population every year. In Europe, mental disorders lead to higher disease costs than chronic physical disorders such as cancer, diabetes or cardiovascular diseases.1,2 In Germany, about every third woman and every fourth man are affected by mental disorders.3 More than 40% of ‘Years lived with disability’ of all disease groups are due to mental illness.2 Employers lose 27–35 working days per year per employee suffering from depression.4 In 2015 in Germany the number of new health-related early retirees was 173 000, 43% of them due to a mental disorder and only 12% due to musculoskeletal disorders.5 Several studies have found that low social support is associated with health impairments and job losses, and is predictive of depressive disorders2,6 and burnout.7 Social support has a moderating effect on workplace psychosocial stressors and their deleterious consequences.8 It is also associated with the state of health and well-being of workers.8 High social support is known as a major factor in coping with stress, disease9 and lower rates of morbidity and mortality.10 In the body of literature there are complex and heterogeneous concepts of social support and different methods of assessing it, each depending on research interest.11 Different contents, sources, network levels and occasions of social support were described and distinguished. Other research differentiates perceived and sustained support or the need and the search for social support.12–16 First of all, a person must recognize his or her support needs to perceive, request and mobilize support.17 This process is determined by personality traits and the environment. Usually two alternative causal models are applied to explain how social support affects psychological distress, namely the direct effect model and the indirect (buffer) effect model.18 The direct effect model indicates that social support enhances well-being irrespective of peoples’ stress levels. The second model proposes support to be mainly related to well-being in stressful stages of peoples’ life.19 Cohort studies have shown that social support from colleagues and supervisors is an important prognostic factor for return to workplace, such as in back pain patients or musculoskeletal problems.20,21 However, currently there are no reliable findings on the impact of subjectively perceived social support outside the workplace on work ability. To our knowledge, the importance of concrete social support in the workplace e.g. by supervisors has been studied but directly effect of social support outside the work on work ability has been hardly investigated. The balance of individual physical, mental and social abilities on the one hand and specific conditions and requirements of the work on the other hand is described as work ability.22 Thus, work ability means the performance of a person regarding his or her skills, abilities and properties in an particular work situation, to successfully meet the tasks required. Work ability is determined by individual potentials as well as internal and external conditions. Since the 1980s, work ability of workers has been assessed by means of the work ability index (WAI) developed in Finland. The WAI assesses how a worker can cope his or her job demands considering his or her health state. In cohort studies, social support has been hardly explored as a potential resource for maintaining work ability and occupational participation. The aim of this study is to examine the importance of perceived social support for self-reported mental health, seeking professional care due to mental health problems and work ability. Well-known relevant and potentially confounders will be considered. Methods Sample and recruitment Data came from the first two waves of the Third German Socio-Medical Panel of Employees (GSPE-III), an ongoing cohort study.23 The first survey was performed in 2013, follow-up data were assessed in 2015. The sample of 10 000 persons (5000 men, 5000 women) was randomly drawn from the register of the German Pension Insurance (GPI). Sampling was restricted to employees aged 40–54 years who were sick for more than six weeks and therefore received sickness absence benefits in 2012. The cohort thus covers vulnerable employees with an increased risk of early retirement or job loss due to health issues. Those with a previous pension request were excluded as well as individuals who had requested or used rehabilitation services during the last four years. Individuals with low mental health at baseline were not excluded from the study. The present analyses included only self-reported data from employed persons participating in both surveys. In each wave, the first questionnaire was followed by a reminder after six weeks. About 89.0% of the employed participants of the baseline survey gave their consent for further follow-up surveys. The study protocol was approved by the ethics committee of the Hannover Medical School (ethics vote no. 1730–2013) and the data protection commissioner of the Federal GPI. The GSPE-III was registered in the German Clinical Trials Register (DRHS no. 00004824). Measures Social support was measured using the Oslo-3 Social Support Scale, an inventory used in Europe to measure perceived social support.24 The three items on the scale capture the number of associated persons in case of serious problems (1: none, 2: 1–2 persons, 3: 3–4 persons, 4: ≥5 persons), the amount of interest and the participation of other people (1: a lot, 2: some, 3: uncertain, 4: little, 5: no), as well as the convenience of the neighbourhood in providing practical assistance when needed (1: very, 2: difficult, 3: possible, 4: easy, 5: very easy). The items were summarized (3–14 points) and the scores were categorized as low (3–8), moderate (9–11) or high social support (12–14).24 Mental health was recorded with the corresponding five-item scale of the Short Form 36-Item Health Survey (SF-36v1, four-week time window).25 The SF-36 is an internationally recognized and widely used generic, validated measure to capture the health-related quality of life of a person. The five-item scale for mental health comprises general mental health, depression, anxiety, emotional and behavioural control as well as general positive mood. The score of the five-item scale ranges from 0 to 100 points. Higher values represent better mental health. In addition, the participants were asked if a psychologist was consulted (0: no, 1: yes). Work ability was assessed with WAI comprising 10 items: (i) current work ability compared with the lifetime best, i.e. the best work ability that has ever been experienced by the participant; (ii) work ability in relation to physical and mental demands of the job (two items); (iii) number of current diseases diagnosed by a physician; (iv) estimated work impairment due to diseases; (v) sick leave during the past year; (vi) own prognosis of work ability two years from now and (vii) mental resources (three items).25 It indicates how a person is able to cope with his or her job demands, taking his or her health into account.26 Values range from 7 to 49 points. Higher values represent better work ability. Additional covariates were socio-demographic characteristics such as age (≥50 years vs. <50 years), sex, life partners (no vs. yes) and educational level (lower secondary level or less vs. upper secondary level). As an indicator of unjust work and psychosocial workload, the ratio of effort and reward (effort–reward ratio) was used. Effort was assessed by three items, reward by seven items. Effort was weighted so that a ratio >1 indicates an imbalance of effort and reward.27 As a further occupational risk factor, the physical job demands were assessed by five items.28 The total score ranged from 0 to 15 points. Higher values represent higher physical job demands. Personality factors were measured with the 15-item Big Five Inventory.29 Each of the five scales (neuroticism, extraversion, openness for new experience, agreeableness and conscientiousness) consisted of three seven-point scaled items. The total scores ranged from 3 to 21 points. Higher values represent a stronger expression of the corresponding personality dimension. Data analyses The sample was characterized using descriptive statistics. Associations between social support and mental health (Model 1) and social support and subjective work ability score (Model 2) were determined using linear regression. Logistic regressions were calculated to clarify the effect of social support on consulting a psychologist (Model 3). As an effect estimate, unstandardized regression coefficients (b) for the linear regressions and odds ratios for the logistic regression (including 95% confidence intervals) were calculated. In a first step, the association of social support with the dependent variables was calculated by adjusting for baseline values of the dependent variables and for socio-demographic characteristics so that known confounders were considered. In a second step, we additionally adjusted for work-related factors and personality traits. This step-by-step approach was used to clarify the influence of work and personality factors. High social support was used as the reference category in regression analyses. Only those persons who were employed at the time of both surveys were included in the analyses. Missing data was imputed by multiple chained equations. Five datasets were generated. The parameter estimators were combined according to Rubin’s rules,30 i.e. coefficients were first estimated in the five completed datasets and afterwards averaged. Differences were considered significant if the double-sided error probability was less than 5%. The analyses were carried out with SPSS 22.0 (IBM, Armonk, USA). Imputation of missing data was done with STATA SE 13 (StataCorp, College Station, USA). Results Sample Figure 1 shows the participants’ flow: a random sample of 10 000 persons was drawn, of whom 103 could not be contacted And 3294 completed the baseline questionnaire (net response rate 33.3%). A comparison showed only marginal differences between participants (mean age 47.9% years, 53.6% women) and non-participants (mean age 47.2% years, 48.4% women). Figure 1 View largeDownload slide Flow of participants Figure 1 View largeDownload slide Flow of participants We excluded unemployed people and people with missing data on employment (n = 311). Of 2983 remaining cases, 328 people did not consent to a follow-up survey (11.0% of employed responders), 559 did not respond and 36 could not be reached. Furthermore, 174 were unemployed or had missing data in employment status. As a result, the analysis sample reduced to 1886 participants. Participants were 48.0 (SD = 4.0) years old on average, 56.5% were female and 28.7% of the participants perceived low social support. Mean self-rated mental health at baseline was 66.3 (SD = 20.1) and 65.8 (SD = 20.0) at follow-up, the WAI was 34.2 (SD = 7.6) and 34.7 (SD = 8.2), respectively. The proportion of persons who had visited a psychologist during the last 12 months decreased from 21.3 to 16.0%. Table 1 depicts characteristics of sample in detail. Table 1 Characteristics of the analysis sample (employed responders; n=1886) Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Table 1 Characteristics of the analysis sample (employed responders; n=1886) Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Social support as a predictor of mental health Model 1 shows a significant negative association of low and moderate social support with mental health after two years (table 2). Adjusted for socio-demographic characteristics, low social support was negatively associated with mental health (Model 1 Step 1: b = −5.59; 95% CI: −7.69 to −3.49). After work-related factors and personality traits were additionally included (Model 1 Step 2: b = −4.96; 95% CI: −7.11 to −2.81), this association was slightly reduced. Table 2 Results of linear and logistic regressions Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  n=1886; 5-fold imputed data; b =unstandardized parameter estimates; CI =confidence interval; OR=odds ratio; Step 1: adjusted for baseline values of the dependent variables and for socio-demographic characteristics; Step 2: additionally adjusted for work-related factors and personality traits. Table 2 Results of linear and logistic regressions Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  n=1886; 5-fold imputed data; b =unstandardized parameter estimates; CI =confidence interval; OR=odds ratio; Step 1: adjusted for baseline values of the dependent variables and for socio-demographic characteristics; Step 2: additionally adjusted for work-related factors and personality traits. Social support as a predictor of the work ability index Adjusted for socio-demographic factors, low social support was slightly negatively associated with follow-up work ability (table 2, Model 2 Step 1: b = −1.22, 95% CI −2.07 to −0.37). Further adjustment of work-related factors and personality traits reduced the effect estimate only marginally (Model 2 Step 2: b = −1.10, 95% CI −2.00 to −0.21). Moderate social support did not affect follow-up work ability. Social support as a predictor for consulting a psychologist In Model 3 Step 1, adjusting for socio-demographic characteristics, low social support increased the odds for consulting a psychologist during the follow-up period (OR = 1.57; 95% CI: 1.08–2.29). After further adjusting for work and personality traits low social support did not significantly increase the likelihood of consulting a psychologist (Model 3 Step 2: OR = 1.30; 95% CI: 0.86–1.96). Discussion The importance of perceived social support for self-rated mental health, work ability and the incidence of a psychological consultation were examined. Potentially influencing factors such as gender, age, partnership, education, effort-reward imbalance, physical workload and personality traits were considered. Compared to a representative sample of Germans of the same age, study participants reported comparatively low values for self-reported mental health25 and moderate values for work ability.26 A large proportion of our participants (28.7%) reported low social support. Our analyses confirmed that low perceived social support is significantly associated with lower mental health and lower subjective work ability. These associations were significant even when adjusting for socio-demographic characteristics, work-related factors and personality traits. Thus, perceived social support can be considered a predictor independent of known risk factors for health and occupational characteristics. Similar results using the Oslo-3 Social Support Scale are obtained from Asian cross-sectional studies which showed significantly better work ability to be associated with better social support.31 A Swedish study showed that women with a long period of incapacity to work assessed their ability to work as higher if there were supporting conditions such as a positive style of leadership and social support. Compared to women without supportive conditions, they also were more likely to return to work.32 Our results are in line with recent state of research concerning the importance of perceived social support on mental health.33–35 In this study, low social support—adjusted for baseline values of the dependent variables and for socio-demographic characteristics—increased the likelihood of consulting a psychologist (Model 3 Step 1). However, this association was not statistically significant if work-related factors and personality traits are considered as covariates (Model 3 Step 2). Thus, unfavourable work-related factors and personality traits might complicate access to psychological consulting. Regrettably, just these individuals have a low odds ratio for mental health and need more consulting. Adjusting for further factors, e. g. level of information or reachability of psychologist was not possible with the available data. In Germany, payment or reimbursement of psychological consulting is completely covered by health insurance. Nevertheless, those suffering from limited social support might need more help to get access to therapeutic support as they may have insufficient problem-solving activities. Strengths of the study are the prospective design, the large sample and the usage of validated instruments. However, the following limitations need to be considered when interpreting the results. First, only persons who are covered by the federal pension insurance scheme and were ill in 2012 for at least six weeks were included. Sampling was focused on employees aged 40–54 years because an age of 40 or more years is associated with an increasing number of musculoskeletal36 and mental disorders.37 Second, the response rate of the first survey was only about one-third. This was comparable to the average response rate, which was meta-analytically derived from past decades38 and is consistent with the observed decline in willingness to participate in surveys in Germany.39,40 We assume that a potential selection bias might implicate the risk of underestimation of the prevalence of low social support and an overestimation of mental health because lower socioeconomic groups are underrepresented in epidemiological studies.41 In our study, a comparison of participants and non-participants showed only marginal differences related to age and sex. Third, after follow-up and applying all exclusion criteria, the analysis sample comprised only 1886 individuals, i.e. only almost 19%. As a consequence, a second selection bias may have been induced by differential dropout. Fortunately, a higher dropout of persons with poor low social support at baseline does not affect the odds ratio regarding low social support and poor follow-up health outcomes. The same is true for a higher dropout of persons with poor follow-up health outcomes. Longitudinal associations of low social support and our outcomes will be underestimated only in case of a higher dropout of persons combining both properties. Fourth, the Oslo-3 Social Support Scale is an instrument which was shown to predict anxiety, depression, mental health and quality of life.42 Differences in social support are known in terms of sex, age, family status, family size, country, socioeconomic and migratory status.42 Therefore, an attempt was made to take the corresponding context factors into account which were available. In addition, personality factors were included. These may influence the behaviour, such as coping strategies, thereby reinforcing the impediments to return to work.43 Due to acceptance considerations, the Oslo-3 Social Support Scale was used as a short, validated questionnaire. Thus, only statements about perceived social support were evaluated, i.e. the subjective perception and interpretation of the participants. So far our conceptualization is restricted to direct relationships of subjective perceived support measured with the Oslo-3 Social Support Scale. Fifth, we analyzed the direct influence of social support on the outcomes. It remains open to what extent social support also acts as a mediator or a moderator for somatic impairments on psychological stress. In a Norwegian Study of older adults, social support acted as a mediator, implying that the negative effect of somatic health problems on psychological distress was mediated by low social support.44 Also, in an Australian longitudinal study, social support served as a buffer for poor mental health after invalidity.44 There is evidence that social support modifies the association between disability and mental health. The largest effects appear for those experiencing a change from high to low social support subsequent to disability and for people with consistently low social support.44 We plan to address this issue using the GSPE-III data in further analyses. To conclude, the interrelations found between perceived social support and mental health or work ability suggest that social support should be given more weight when it comes to restoring or maintaining mental health and work ability. In the body of literature, elements of external conditions such as social support outside of work or the current labour market situation might be insufficiently considered. Our results support research findings which, in the context of work ability, call for the need to expand the prospects for the compatibility of work and private life.23 Social relations that can be supportive must be nurtured. Focusing on work to the detriment of social relations involves risks to the preservation of work ability. In stressful situations, family members, friends, neighbours and work colleagues can become an important resource in order to navigate through the health system and to make use of existing services. This is exactly why perceived social support should be considered and may serve as a supplementing, valuable predictor in getting an overall idea of work ability. Funding The study is funded by Federation of German Pension Insurance Institutions (8011 – 106 – 31/31.104.1). Disclaimer The authors declare that they have no financial or non-financial competing interests. Conflicts of interest: None declared. Key points Based on results of our cohort study, low social support is a risk factor for mental health and work ability two years later. Social support also outside work should be given more weight when planning strategies in order to restore or maintain work ability. Low social support was not associated with consulting a psychologist in the fully adjusted model. References 1 Gustavsson A, Svensson M, Jacobi F, et al.   Cost of disorders of the brain in Europe 2010. Eur Neuropsychopharmacol  2011; 21: 718– 79. Google Scholar CrossRef Search ADS PubMed  2 Wittchen HU, Jacobi F, Rehm J, et al.   The size and burden of mental disorders and other disorders of the brain in Europe 2010. Eur Neuropsychopharmacol  2011; 21: 655– 79. Google Scholar CrossRef Search ADS PubMed  3 Jacobi F, Hofler M, Strehle J, et al.   Psychische Störungen in der Allgemeinbevölkerung: studie zur Gesundheit Erwachsener in Deutschland und ihr Zusatzmodul Psychische Gesundheit (DEGS1-MH) (Mental disorders in the general population: study on the health of adults in Germany and the additional module mental health (DEGS1-MH)). Nervenarzt  2014; 85: 77– 87. Google Scholar CrossRef Search ADS PubMed  4 World Health Organization. Preventing depression in the WHO European Region, Copenhagen, 2016. 5 Deutsche Rentenversicherung. Rentenversicherung in Zeitreihen (Pension Insurance in Time Series) . Berlin: Deutsche Rentenversicherung Bund, 2016. 6 Niedhammer I, Malard L, Chastang JF. Occupational factors and subsequent major depressive and generalized anxiety disorders in the prospective French national SIP study. BMC Public Health  2015; 15: 200. Google Scholar CrossRef Search ADS PubMed  7 Adriaenssens J, De Gucht V, Maes S. Determinants and prevalence of burnout in emergency nurses: a systematic review of 25 years of research. Int J Nurs Stud  2015; 52: 649– 61. Google Scholar CrossRef Search ADS PubMed  8 Sargent LD, Terry DJ. The moderating role of social support in Karasek’s job strain model. Work Stress  2000; 14: 245– 61. Google Scholar CrossRef Search ADS   9 Kroll LE, Lampert T. Unemployment, social support and health problems: results of the GEDA study in Germany, 2009. Dtsch Arztebl Int  2011; 108: 47– 52. Google Scholar PubMed  10 Uchino BN. Social support and health: a review of physiological processes potentially underlying links to disease outcomes. Behav Sci  2006; 29: 377– 87. 11 Eisele M, Zimmermann T, Köhler M, et al.   Influence of social support on cognitive change and mortality in old age: results from the prospective multicentre cohort study AgeCoDe. BMC Geriatr  2012; 12: 9. Google Scholar CrossRef Search ADS PubMed  12 Cyranowski JM, Zill N, Bode R, et al.   Assessing social support, companionship, and distress: National Institute of Health (NIH) Toolbox Adult Social Relationship Scales. Health Psychol  2013; 32: 293– 301. 13 Woodward AT, Taylor RJ, Neighbors HW, et al.   Use of professional and informal support by African Americans and Caribbean blacks with mental disorders. Psychiatr Serv  2008; 59: 1292– 8. Google Scholar CrossRef Search ADS PubMed  14 Berkman LF, Glass T. Social integration, social networks, social support and health. In: Berkman LF, Kawachi I, editors. Social Epidemiology . New York: Oxford University Press, 2000: 137– 73. 15 Schwarzer R, Leppin A. Social support and health: a theoretical and empirical overview. J Soc Pers Relat  1991; 8: 99– 127. Google Scholar CrossRef Search ADS   16 Tough H, Siegrist J, Fekete C. Social relationships, mental health and wellbeing in physical disability: a systematic review. BMC Public Health  2017; 17: 414. Google Scholar CrossRef Search ADS PubMed  17 Schwarzer R, Knoll N, Rieckmann N. Social support. In: Kaptein A, Weinman J, editors. Health Psychology . Oxford, England: Blackwell, 2004: 158– 82. 18 Cohen S, Wills TA. Stress, social support, and the buffering hypothesis. Psychol Bull  1985; 98: 310– 57. Google Scholar CrossRef Search ADS PubMed  19 Rodriguez M, Cohen S. Social support. In: Friedman H, editor. Encyclopedia of Mental Health . New York: Academis Press, 1998: 535– 44. 20 Bethge M. Rückenschmerzpatienten (Patients suffering from back pain). Orthopade  2010; 39: 866– 73. Google Scholar CrossRef Search ADS PubMed  21 Woods V. Work-related musculoskeletal health and social support. Occup Med  2005; 55: 177– 89. Google Scholar CrossRef Search ADS   22 Ilmarinen J. Work ability – a comprehensive concept for occupational health research and prevention. Scand J Work Environ Health  2009; 35: 1– 5. Google Scholar CrossRef Search ADS PubMed  23 Bethge M, Spanier K, Neugebauer T, et al.   Self-reported poor work ability – an indicator of need for rehabilitation? A cross-sectional study of a sample of German employees. Am J Phys Med Rehabil  2015; 94: 958– 66. Google Scholar CrossRef Search ADS PubMed  24 Dalgard OS, Dowrick C, Lehtinen V, et al.   Negative life events, social support and gender difference in depression: a multinational community survey with data from the ODIN study. Soc Psychiatry Psychiatr Epidemiol  2006; 41: 444. Google Scholar CrossRef Search ADS PubMed  25 Bullinger M, Kirchberger I. SF-36 Fragebogen Zum Gesundheitszustand: Handanweisung (SF-36 Health Survey Questionnaire: Hand Statement) . Göttingen: Hogrefe, 1998. 26 Tuomi K, Ilmarinen J, Jahkola A, et al.   Work Ability Index , 2nd rev edn Helsinki: Finnish Institute of Occupational Health, 1998. 27 Siegrist J, Wege N, Puhlhofer F, Wahrendorf M. A short generic measure of work stress in the era of globalization: effort-reward imbalance. Int Arch Occup Environ Health  2009; 82: 1005– 13. Google Scholar CrossRef Search ADS PubMed  28 Slesina W. Arbeitsbedingte Erkrankungen Und Arbeitsanalyse: Arbeitsanalyse Unter Dem Gesichtspunkt Der Gesundheitsvorsorge (Work-Related Diseases and Work Analysis: Analysis of Work in Terms of Health Care) . Stuttgart: Enke, 1987. 29 Lang FR, John D, Lüdtke O, et al.   Short assessment of the Big Five: robust across survey methods except telephone interviewing. Behav Res Methods  2011; 43: 548– 67. Google Scholar CrossRef Search ADS PubMed  30 Little RJA, Rubin DB. Statistical Analysis with Missing Data . Hoboken: Wiley, 2002. Google Scholar CrossRef Search ADS   31 Han L, Shi L, Lu L, Ling L. Work ability of Chinese migrant workers: the influence of migration characteristics. BMC Public Health  2014; 14: 353. Google Scholar CrossRef Search ADS PubMed  32 Ahlstrom L, Hagberg M, Dellve L. Workplace rehabilitation and supportive conditions at work: a prospective study. J Occup Rehabil  2013; 23: 248– 60. Google Scholar CrossRef Search ADS PubMed  33 Vemer P, Bouwmans CA, Zijlstra-Vlasveld MC, et al.   Let's get back to work: survival analysis on the return-to-work after depression. Neuropsychiatr Dis Treat  2013; 9: 1637– 45. Google Scholar PubMed  34 van Hees ML, Rotter T, Ellermann T, Evers SM. The effectiveness of individual interpersonal psychotherapy as a treatment for major depressive disorder in adult outpatients: a systematic review. BMC Psychiatry  2013; 13: 22. Google Scholar CrossRef Search ADS PubMed  35 Uchino BN, Birmingham W. Stess and support processes. In: Contrado RJ, Baum A, editors. The Handbook of Stress Science: Biology, Psychology, and Health . New York: Springer Publishing Co, 2010: 111– 21. 36 The burden of musculoskeletal disease–a global perspective. Brooks PM. Clin Rheumatol  2006; 25: 778– 81. Epub 2006 Apr 12. Review. CrossRef Search ADS PubMed  37 Kessler RC, Angermeyer M, Anthony JC, et al.   Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization's World Mental Health Survey Initiative. World Psychiatry  2007; 6: 168– 76. Google Scholar PubMed  38 Aust F, Schröder H. Sinkende Stichprobenausschöpfung in der Umfrageforschung: ein Bericht aus der Praxis (Declining sampling coverage in survey research: a report from practical experience). OZS Osterr Z Soziol  2009; 34: 195– 212. 39 Schnell R. Survey-Interviews: Methoden Standardisierter Befragungen (Survey Interviews: Methods of Standardised Assessments) . Wiesbaden: Springer VS, 2012. 40 Groves RM. Survey Errors and Survey Costs . New York: Wiley, 1989. Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The European Journal of Public Health Oxford University Press

Influence of social support among employees on mental health and work ability—a prospective cohort study in 2013–15

Loading next page...
 
/lp/ou_press/influence-of-social-support-among-employees-on-mental-health-and-work-D5NioUxw8E
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.
ISSN
1101-1262
eISSN
1464-360X
D.O.I.
10.1093/eurpub/cky067
Publisher site
See Article on Publisher Site

Abstract

Abstract Background The study examined the association of social support with mental health, subjective work ability and psychological consultation. Methods The cohort study included 1886 German employees (40–54 years) with sickness absence exceeding six weeks in 2012. Postal surveys were carried out in 2013 and 2015. Results After adjustment for socio-demographic data, work-related characteristics and personality factors, persons with low social support compared to high social support had poorer mental health (b=−4.96; 95% CI: −7.11 to −2.81) and a lower work ability index (b=−1.10; 95% CI −2.00 to −0.21). Low social support was not associated with increased odds of consulting a psychologist (OR =1.30; 95% CI: 0.86–1.96). Conclusions Low social support is an independent predictor of poorer mental health and lower work ability. This study highlights the importance of identifying people who have limited access to social support. Introduction In Europe, there exists a great demand for effective measures to improve work ability among those threatened by illness. In recent years this applies particularly to mental disorders.1,2 Mental well-being is a key resource for productivity, participation and inclusion. Mental disorders are among the top ten public health challenges in the WHO European Region, affecting about 25% of the population every year. In Europe, mental disorders lead to higher disease costs than chronic physical disorders such as cancer, diabetes or cardiovascular diseases.1,2 In Germany, about every third woman and every fourth man are affected by mental disorders.3 More than 40% of ‘Years lived with disability’ of all disease groups are due to mental illness.2 Employers lose 27–35 working days per year per employee suffering from depression.4 In 2015 in Germany the number of new health-related early retirees was 173 000, 43% of them due to a mental disorder and only 12% due to musculoskeletal disorders.5 Several studies have found that low social support is associated with health impairments and job losses, and is predictive of depressive disorders2,6 and burnout.7 Social support has a moderating effect on workplace psychosocial stressors and their deleterious consequences.8 It is also associated with the state of health and well-being of workers.8 High social support is known as a major factor in coping with stress, disease9 and lower rates of morbidity and mortality.10 In the body of literature there are complex and heterogeneous concepts of social support and different methods of assessing it, each depending on research interest.11 Different contents, sources, network levels and occasions of social support were described and distinguished. Other research differentiates perceived and sustained support or the need and the search for social support.12–16 First of all, a person must recognize his or her support needs to perceive, request and mobilize support.17 This process is determined by personality traits and the environment. Usually two alternative causal models are applied to explain how social support affects psychological distress, namely the direct effect model and the indirect (buffer) effect model.18 The direct effect model indicates that social support enhances well-being irrespective of peoples’ stress levels. The second model proposes support to be mainly related to well-being in stressful stages of peoples’ life.19 Cohort studies have shown that social support from colleagues and supervisors is an important prognostic factor for return to workplace, such as in back pain patients or musculoskeletal problems.20,21 However, currently there are no reliable findings on the impact of subjectively perceived social support outside the workplace on work ability. To our knowledge, the importance of concrete social support in the workplace e.g. by supervisors has been studied but directly effect of social support outside the work on work ability has been hardly investigated. The balance of individual physical, mental and social abilities on the one hand and specific conditions and requirements of the work on the other hand is described as work ability.22 Thus, work ability means the performance of a person regarding his or her skills, abilities and properties in an particular work situation, to successfully meet the tasks required. Work ability is determined by individual potentials as well as internal and external conditions. Since the 1980s, work ability of workers has been assessed by means of the work ability index (WAI) developed in Finland. The WAI assesses how a worker can cope his or her job demands considering his or her health state. In cohort studies, social support has been hardly explored as a potential resource for maintaining work ability and occupational participation. The aim of this study is to examine the importance of perceived social support for self-reported mental health, seeking professional care due to mental health problems and work ability. Well-known relevant and potentially confounders will be considered. Methods Sample and recruitment Data came from the first two waves of the Third German Socio-Medical Panel of Employees (GSPE-III), an ongoing cohort study.23 The first survey was performed in 2013, follow-up data were assessed in 2015. The sample of 10 000 persons (5000 men, 5000 women) was randomly drawn from the register of the German Pension Insurance (GPI). Sampling was restricted to employees aged 40–54 years who were sick for more than six weeks and therefore received sickness absence benefits in 2012. The cohort thus covers vulnerable employees with an increased risk of early retirement or job loss due to health issues. Those with a previous pension request were excluded as well as individuals who had requested or used rehabilitation services during the last four years. Individuals with low mental health at baseline were not excluded from the study. The present analyses included only self-reported data from employed persons participating in both surveys. In each wave, the first questionnaire was followed by a reminder after six weeks. About 89.0% of the employed participants of the baseline survey gave their consent for further follow-up surveys. The study protocol was approved by the ethics committee of the Hannover Medical School (ethics vote no. 1730–2013) and the data protection commissioner of the Federal GPI. The GSPE-III was registered in the German Clinical Trials Register (DRHS no. 00004824). Measures Social support was measured using the Oslo-3 Social Support Scale, an inventory used in Europe to measure perceived social support.24 The three items on the scale capture the number of associated persons in case of serious problems (1: none, 2: 1–2 persons, 3: 3–4 persons, 4: ≥5 persons), the amount of interest and the participation of other people (1: a lot, 2: some, 3: uncertain, 4: little, 5: no), as well as the convenience of the neighbourhood in providing practical assistance when needed (1: very, 2: difficult, 3: possible, 4: easy, 5: very easy). The items were summarized (3–14 points) and the scores were categorized as low (3–8), moderate (9–11) or high social support (12–14).24 Mental health was recorded with the corresponding five-item scale of the Short Form 36-Item Health Survey (SF-36v1, four-week time window).25 The SF-36 is an internationally recognized and widely used generic, validated measure to capture the health-related quality of life of a person. The five-item scale for mental health comprises general mental health, depression, anxiety, emotional and behavioural control as well as general positive mood. The score of the five-item scale ranges from 0 to 100 points. Higher values represent better mental health. In addition, the participants were asked if a psychologist was consulted (0: no, 1: yes). Work ability was assessed with WAI comprising 10 items: (i) current work ability compared with the lifetime best, i.e. the best work ability that has ever been experienced by the participant; (ii) work ability in relation to physical and mental demands of the job (two items); (iii) number of current diseases diagnosed by a physician; (iv) estimated work impairment due to diseases; (v) sick leave during the past year; (vi) own prognosis of work ability two years from now and (vii) mental resources (three items).25 It indicates how a person is able to cope with his or her job demands, taking his or her health into account.26 Values range from 7 to 49 points. Higher values represent better work ability. Additional covariates were socio-demographic characteristics such as age (≥50 years vs. <50 years), sex, life partners (no vs. yes) and educational level (lower secondary level or less vs. upper secondary level). As an indicator of unjust work and psychosocial workload, the ratio of effort and reward (effort–reward ratio) was used. Effort was assessed by three items, reward by seven items. Effort was weighted so that a ratio >1 indicates an imbalance of effort and reward.27 As a further occupational risk factor, the physical job demands were assessed by five items.28 The total score ranged from 0 to 15 points. Higher values represent higher physical job demands. Personality factors were measured with the 15-item Big Five Inventory.29 Each of the five scales (neuroticism, extraversion, openness for new experience, agreeableness and conscientiousness) consisted of three seven-point scaled items. The total scores ranged from 3 to 21 points. Higher values represent a stronger expression of the corresponding personality dimension. Data analyses The sample was characterized using descriptive statistics. Associations between social support and mental health (Model 1) and social support and subjective work ability score (Model 2) were determined using linear regression. Logistic regressions were calculated to clarify the effect of social support on consulting a psychologist (Model 3). As an effect estimate, unstandardized regression coefficients (b) for the linear regressions and odds ratios for the logistic regression (including 95% confidence intervals) were calculated. In a first step, the association of social support with the dependent variables was calculated by adjusting for baseline values of the dependent variables and for socio-demographic characteristics so that known confounders were considered. In a second step, we additionally adjusted for work-related factors and personality traits. This step-by-step approach was used to clarify the influence of work and personality factors. High social support was used as the reference category in regression analyses. Only those persons who were employed at the time of both surveys were included in the analyses. Missing data was imputed by multiple chained equations. Five datasets were generated. The parameter estimators were combined according to Rubin’s rules,30 i.e. coefficients were first estimated in the five completed datasets and afterwards averaged. Differences were considered significant if the double-sided error probability was less than 5%. The analyses were carried out with SPSS 22.0 (IBM, Armonk, USA). Imputation of missing data was done with STATA SE 13 (StataCorp, College Station, USA). Results Sample Figure 1 shows the participants’ flow: a random sample of 10 000 persons was drawn, of whom 103 could not be contacted And 3294 completed the baseline questionnaire (net response rate 33.3%). A comparison showed only marginal differences between participants (mean age 47.9% years, 53.6% women) and non-participants (mean age 47.2% years, 48.4% women). Figure 1 View largeDownload slide Flow of participants Figure 1 View largeDownload slide Flow of participants We excluded unemployed people and people with missing data on employment (n = 311). Of 2983 remaining cases, 328 people did not consent to a follow-up survey (11.0% of employed responders), 559 did not respond and 36 could not be reached. Furthermore, 174 were unemployed or had missing data in employment status. As a result, the analysis sample reduced to 1886 participants. Participants were 48.0 (SD = 4.0) years old on average, 56.5% were female and 28.7% of the participants perceived low social support. Mean self-rated mental health at baseline was 66.3 (SD = 20.1) and 65.8 (SD = 20.0) at follow-up, the WAI was 34.2 (SD = 7.6) and 34.7 (SD = 8.2), respectively. The proportion of persons who had visited a psychologist during the last 12 months decreased from 21.3 to 16.0%. Table 1 depicts characteristics of sample in detail. Table 1 Characteristics of the analysis sample (employed responders; n=1886) Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Table 1 Characteristics of the analysis sample (employed responders; n=1886) Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Variable  Category  n  Mean (SD) or %   Baseline        Age  %≥50  1886  40.2  Sex  % female  1886  56.5  Life partner  % no  1850  20.5  Education level  % low  1880  16.5  Effort-reward ratio  –  1829  1.4 (0.6)  Physical job demands  –  1848  4.7 (4.9)  Big 5: neuroticism  3–21  1871  12.3 (4.1)      Extraversion  3–21  1873  14.8 (3.5)      Openness  3–21  1860  14.4 (3.5)      Conscientiousness  3–21  1873  18.6 (2.3)      Agreeableness  3–21  1876  16.7 (2.9)  Oslo 3 social support scale  % low (3–8 points)  1872  28.7  % moderate (9–11 points)  1872  48.0  % high (12–14 points)  1872  23.3  SF-36: mental health  0–100  1877  66.3 (20.1)  Work ability  7–49  1807  34.2 (7.6)  Psychological consultation  yes  1886  21.3  Follow-up        SF-36: mental health  0–100  1868  65.8 (20.0)  Work ability  7–49  1824  34.7 (8.2)  Psychological consultation  yes  1886  16.0  Social support as a predictor of mental health Model 1 shows a significant negative association of low and moderate social support with mental health after two years (table 2). Adjusted for socio-demographic characteristics, low social support was negatively associated with mental health (Model 1 Step 1: b = −5.59; 95% CI: −7.69 to −3.49). After work-related factors and personality traits were additionally included (Model 1 Step 2: b = −4.96; 95% CI: −7.11 to −2.81), this association was slightly reduced. Table 2 Results of linear and logistic regressions Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  n=1886; 5-fold imputed data; b =unstandardized parameter estimates; CI =confidence interval; OR=odds ratio; Step 1: adjusted for baseline values of the dependent variables and for socio-demographic characteristics; Step 2: additionally adjusted for work-related factors and personality traits. Table 2 Results of linear and logistic regressions Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  Model 1: Mental health  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Mental health at baseline, 0–100  0.59 (0.56; 0.63)  0.51 (0.46; 0.55)  Social support, low vs. high  −5.59 (−7.69; −3.49)  −4.96 (−7.11; −2.81)  Social support, moderate vs. high  −2.50 (−4.30; −0.70)  −1.93 (−3.73; 0.13)  Model 2: Work ability index  Step 1  Step 2  Predictors  b (95% CI)  b (95% CI)  Work ability index at baseline, 7–49  0.64 (0.60; 0.68)  0.59 (0.55; 0.64)  Social support, low vs. high  −1.22 (−2.07; −0.37)  −1.10 (−2.00; −0.21)  Social support, moderate vs. high  −0.13 (−0.92; 0.67)  −0.02 (−0.82; 0.77)  Model 3: Psychological consultation  Step 1  Step 2  Predictors  OR (95% CI)  OR (95% CI)  Psychological consultation at baseline, no vs. yes  6.62 (5.05; 8.69)  5.50 (4.13; 7.33)  Social support, low vs. high  1.57 (1.08; 2.29)  1.30 (0.86; 1.96)  Social support, moderate vs. high  1.09 (0.76; 1.56)  0.99 (0.69; 1.44)  n=1886; 5-fold imputed data; b =unstandardized parameter estimates; CI =confidence interval; OR=odds ratio; Step 1: adjusted for baseline values of the dependent variables and for socio-demographic characteristics; Step 2: additionally adjusted for work-related factors and personality traits. Social support as a predictor of the work ability index Adjusted for socio-demographic factors, low social support was slightly negatively associated with follow-up work ability (table 2, Model 2 Step 1: b = −1.22, 95% CI −2.07 to −0.37). Further adjustment of work-related factors and personality traits reduced the effect estimate only marginally (Model 2 Step 2: b = −1.10, 95% CI −2.00 to −0.21). Moderate social support did not affect follow-up work ability. Social support as a predictor for consulting a psychologist In Model 3 Step 1, adjusting for socio-demographic characteristics, low social support increased the odds for consulting a psychologist during the follow-up period (OR = 1.57; 95% CI: 1.08–2.29). After further adjusting for work and personality traits low social support did not significantly increase the likelihood of consulting a psychologist (Model 3 Step 2: OR = 1.30; 95% CI: 0.86–1.96). Discussion The importance of perceived social support for self-rated mental health, work ability and the incidence of a psychological consultation were examined. Potentially influencing factors such as gender, age, partnership, education, effort-reward imbalance, physical workload and personality traits were considered. Compared to a representative sample of Germans of the same age, study participants reported comparatively low values for self-reported mental health25 and moderate values for work ability.26 A large proportion of our participants (28.7%) reported low social support. Our analyses confirmed that low perceived social support is significantly associated with lower mental health and lower subjective work ability. These associations were significant even when adjusting for socio-demographic characteristics, work-related factors and personality traits. Thus, perceived social support can be considered a predictor independent of known risk factors for health and occupational characteristics. Similar results using the Oslo-3 Social Support Scale are obtained from Asian cross-sectional studies which showed significantly better work ability to be associated with better social support.31 A Swedish study showed that women with a long period of incapacity to work assessed their ability to work as higher if there were supporting conditions such as a positive style of leadership and social support. Compared to women without supportive conditions, they also were more likely to return to work.32 Our results are in line with recent state of research concerning the importance of perceived social support on mental health.33–35 In this study, low social support—adjusted for baseline values of the dependent variables and for socio-demographic characteristics—increased the likelihood of consulting a psychologist (Model 3 Step 1). However, this association was not statistically significant if work-related factors and personality traits are considered as covariates (Model 3 Step 2). Thus, unfavourable work-related factors and personality traits might complicate access to psychological consulting. Regrettably, just these individuals have a low odds ratio for mental health and need more consulting. Adjusting for further factors, e. g. level of information or reachability of psychologist was not possible with the available data. In Germany, payment or reimbursement of psychological consulting is completely covered by health insurance. Nevertheless, those suffering from limited social support might need more help to get access to therapeutic support as they may have insufficient problem-solving activities. Strengths of the study are the prospective design, the large sample and the usage of validated instruments. However, the following limitations need to be considered when interpreting the results. First, only persons who are covered by the federal pension insurance scheme and were ill in 2012 for at least six weeks were included. Sampling was focused on employees aged 40–54 years because an age of 40 or more years is associated with an increasing number of musculoskeletal36 and mental disorders.37 Second, the response rate of the first survey was only about one-third. This was comparable to the average response rate, which was meta-analytically derived from past decades38 and is consistent with the observed decline in willingness to participate in surveys in Germany.39,40 We assume that a potential selection bias might implicate the risk of underestimation of the prevalence of low social support and an overestimation of mental health because lower socioeconomic groups are underrepresented in epidemiological studies.41 In our study, a comparison of participants and non-participants showed only marginal differences related to age and sex. Third, after follow-up and applying all exclusion criteria, the analysis sample comprised only 1886 individuals, i.e. only almost 19%. As a consequence, a second selection bias may have been induced by differential dropout. Fortunately, a higher dropout of persons with poor low social support at baseline does not affect the odds ratio regarding low social support and poor follow-up health outcomes. The same is true for a higher dropout of persons with poor follow-up health outcomes. Longitudinal associations of low social support and our outcomes will be underestimated only in case of a higher dropout of persons combining both properties. Fourth, the Oslo-3 Social Support Scale is an instrument which was shown to predict anxiety, depression, mental health and quality of life.42 Differences in social support are known in terms of sex, age, family status, family size, country, socioeconomic and migratory status.42 Therefore, an attempt was made to take the corresponding context factors into account which were available. In addition, personality factors were included. These may influence the behaviour, such as coping strategies, thereby reinforcing the impediments to return to work.43 Due to acceptance considerations, the Oslo-3 Social Support Scale was used as a short, validated questionnaire. Thus, only statements about perceived social support were evaluated, i.e. the subjective perception and interpretation of the participants. So far our conceptualization is restricted to direct relationships of subjective perceived support measured with the Oslo-3 Social Support Scale. Fifth, we analyzed the direct influence of social support on the outcomes. It remains open to what extent social support also acts as a mediator or a moderator for somatic impairments on psychological stress. In a Norwegian Study of older adults, social support acted as a mediator, implying that the negative effect of somatic health problems on psychological distress was mediated by low social support.44 Also, in an Australian longitudinal study, social support served as a buffer for poor mental health after invalidity.44 There is evidence that social support modifies the association between disability and mental health. The largest effects appear for those experiencing a change from high to low social support subsequent to disability and for people with consistently low social support.44 We plan to address this issue using the GSPE-III data in further analyses. To conclude, the interrelations found between perceived social support and mental health or work ability suggest that social support should be given more weight when it comes to restoring or maintaining mental health and work ability. In the body of literature, elements of external conditions such as social support outside of work or the current labour market situation might be insufficiently considered. Our results support research findings which, in the context of work ability, call for the need to expand the prospects for the compatibility of work and private life.23 Social relations that can be supportive must be nurtured. Focusing on work to the detriment of social relations involves risks to the preservation of work ability. In stressful situations, family members, friends, neighbours and work colleagues can become an important resource in order to navigate through the health system and to make use of existing services. This is exactly why perceived social support should be considered and may serve as a supplementing, valuable predictor in getting an overall idea of work ability. Funding The study is funded by Federation of German Pension Insurance Institutions (8011 – 106 – 31/31.104.1). Disclaimer The authors declare that they have no financial or non-financial competing interests. Conflicts of interest: None declared. Key points Based on results of our cohort study, low social support is a risk factor for mental health and work ability two years later. Social support also outside work should be given more weight when planning strategies in order to restore or maintain work ability. Low social support was not associated with consulting a psychologist in the fully adjusted model. References 1 Gustavsson A, Svensson M, Jacobi F, et al.   Cost of disorders of the brain in Europe 2010. Eur Neuropsychopharmacol  2011; 21: 718– 79. Google Scholar CrossRef Search ADS PubMed  2 Wittchen HU, Jacobi F, Rehm J, et al.   The size and burden of mental disorders and other disorders of the brain in Europe 2010. Eur Neuropsychopharmacol  2011; 21: 655– 79. Google Scholar CrossRef Search ADS PubMed  3 Jacobi F, Hofler M, Strehle J, et al.   Psychische Störungen in der Allgemeinbevölkerung: studie zur Gesundheit Erwachsener in Deutschland und ihr Zusatzmodul Psychische Gesundheit (DEGS1-MH) (Mental disorders in the general population: study on the health of adults in Germany and the additional module mental health (DEGS1-MH)). Nervenarzt  2014; 85: 77– 87. Google Scholar CrossRef Search ADS PubMed  4 World Health Organization. Preventing depression in the WHO European Region, Copenhagen, 2016. 5 Deutsche Rentenversicherung. Rentenversicherung in Zeitreihen (Pension Insurance in Time Series) . Berlin: Deutsche Rentenversicherung Bund, 2016. 6 Niedhammer I, Malard L, Chastang JF. Occupational factors and subsequent major depressive and generalized anxiety disorders in the prospective French national SIP study. BMC Public Health  2015; 15: 200. Google Scholar CrossRef Search ADS PubMed  7 Adriaenssens J, De Gucht V, Maes S. Determinants and prevalence of burnout in emergency nurses: a systematic review of 25 years of research. Int J Nurs Stud  2015; 52: 649– 61. Google Scholar CrossRef Search ADS PubMed  8 Sargent LD, Terry DJ. The moderating role of social support in Karasek’s job strain model. Work Stress  2000; 14: 245– 61. Google Scholar CrossRef Search ADS   9 Kroll LE, Lampert T. Unemployment, social support and health problems: results of the GEDA study in Germany, 2009. Dtsch Arztebl Int  2011; 108: 47– 52. Google Scholar PubMed  10 Uchino BN. Social support and health: a review of physiological processes potentially underlying links to disease outcomes. Behav Sci  2006; 29: 377– 87. 11 Eisele M, Zimmermann T, Köhler M, et al.   Influence of social support on cognitive change and mortality in old age: results from the prospective multicentre cohort study AgeCoDe. BMC Geriatr  2012; 12: 9. Google Scholar CrossRef Search ADS PubMed  12 Cyranowski JM, Zill N, Bode R, et al.   Assessing social support, companionship, and distress: National Institute of Health (NIH) Toolbox Adult Social Relationship Scales. Health Psychol  2013; 32: 293– 301. 13 Woodward AT, Taylor RJ, Neighbors HW, et al.   Use of professional and informal support by African Americans and Caribbean blacks with mental disorders. Psychiatr Serv  2008; 59: 1292– 8. Google Scholar CrossRef Search ADS PubMed  14 Berkman LF, Glass T. Social integration, social networks, social support and health. In: Berkman LF, Kawachi I, editors. Social Epidemiology . New York: Oxford University Press, 2000: 137– 73. 15 Schwarzer R, Leppin A. Social support and health: a theoretical and empirical overview. J Soc Pers Relat  1991; 8: 99– 127. Google Scholar CrossRef Search ADS   16 Tough H, Siegrist J, Fekete C. Social relationships, mental health and wellbeing in physical disability: a systematic review. BMC Public Health  2017; 17: 414. Google Scholar CrossRef Search ADS PubMed  17 Schwarzer R, Knoll N, Rieckmann N. Social support. In: Kaptein A, Weinman J, editors. Health Psychology . Oxford, England: Blackwell, 2004: 158– 82. 18 Cohen S, Wills TA. Stress, social support, and the buffering hypothesis. Psychol Bull  1985; 98: 310– 57. Google Scholar CrossRef Search ADS PubMed  19 Rodriguez M, Cohen S. Social support. In: Friedman H, editor. Encyclopedia of Mental Health . New York: Academis Press, 1998: 535– 44. 20 Bethge M. Rückenschmerzpatienten (Patients suffering from back pain). Orthopade  2010; 39: 866– 73. Google Scholar CrossRef Search ADS PubMed  21 Woods V. Work-related musculoskeletal health and social support. Occup Med  2005; 55: 177– 89. Google Scholar CrossRef Search ADS   22 Ilmarinen J. Work ability – a comprehensive concept for occupational health research and prevention. Scand J Work Environ Health  2009; 35: 1– 5. Google Scholar CrossRef Search ADS PubMed  23 Bethge M, Spanier K, Neugebauer T, et al.   Self-reported poor work ability – an indicator of need for rehabilitation? A cross-sectional study of a sample of German employees. Am J Phys Med Rehabil  2015; 94: 958– 66. Google Scholar CrossRef Search ADS PubMed  24 Dalgard OS, Dowrick C, Lehtinen V, et al.   Negative life events, social support and gender difference in depression: a multinational community survey with data from the ODIN study. Soc Psychiatry Psychiatr Epidemiol  2006; 41: 444. Google Scholar CrossRef Search ADS PubMed  25 Bullinger M, Kirchberger I. SF-36 Fragebogen Zum Gesundheitszustand: Handanweisung (SF-36 Health Survey Questionnaire: Hand Statement) . Göttingen: Hogrefe, 1998. 26 Tuomi K, Ilmarinen J, Jahkola A, et al.   Work Ability Index , 2nd rev edn Helsinki: Finnish Institute of Occupational Health, 1998. 27 Siegrist J, Wege N, Puhlhofer F, Wahrendorf M. A short generic measure of work stress in the era of globalization: effort-reward imbalance. Int Arch Occup Environ Health  2009; 82: 1005– 13. Google Scholar CrossRef Search ADS PubMed  28 Slesina W. Arbeitsbedingte Erkrankungen Und Arbeitsanalyse: Arbeitsanalyse Unter Dem Gesichtspunkt Der Gesundheitsvorsorge (Work-Related Diseases and Work Analysis: Analysis of Work in Terms of Health Care) . Stuttgart: Enke, 1987. 29 Lang FR, John D, Lüdtke O, et al.   Short assessment of the Big Five: robust across survey methods except telephone interviewing. Behav Res Methods  2011; 43: 548– 67. Google Scholar CrossRef Search ADS PubMed  30 Little RJA, Rubin DB. Statistical Analysis with Missing Data . Hoboken: Wiley, 2002. Google Scholar CrossRef Search ADS   31 Han L, Shi L, Lu L, Ling L. Work ability of Chinese migrant workers: the influence of migration characteristics. BMC Public Health  2014; 14: 353. Google Scholar CrossRef Search ADS PubMed  32 Ahlstrom L, Hagberg M, Dellve L. Workplace rehabilitation and supportive conditions at work: a prospective study. J Occup Rehabil  2013; 23: 248– 60. Google Scholar CrossRef Search ADS PubMed  33 Vemer P, Bouwmans CA, Zijlstra-Vlasveld MC, et al.   Let's get back to work: survival analysis on the return-to-work after depression. Neuropsychiatr Dis Treat  2013; 9: 1637– 45. Google Scholar PubMed  34 van Hees ML, Rotter T, Ellermann T, Evers SM. The effectiveness of individual interpersonal psychotherapy as a treatment for major depressive disorder in adult outpatients: a systematic review. BMC Psychiatry  2013; 13: 22. Google Scholar CrossRef Search ADS PubMed  35 Uchino BN, Birmingham W. Stess and support processes. In: Contrado RJ, Baum A, editors. The Handbook of Stress Science: Biology, Psychology, and Health . New York: Springer Publishing Co, 2010: 111– 21. 36 The burden of musculoskeletal disease–a global perspective. Brooks PM. Clin Rheumatol  2006; 25: 778– 81. Epub 2006 Apr 12. Review. CrossRef Search ADS PubMed  37 Kessler RC, Angermeyer M, Anthony JC, et al.   Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization's World Mental Health Survey Initiative. World Psychiatry  2007; 6: 168– 76. Google Scholar PubMed  38 Aust F, Schröder H. Sinkende Stichprobenausschöpfung in der Umfrageforschung: ein Bericht aus der Praxis (Declining sampling coverage in survey research: a report from practical experience). OZS Osterr Z Soziol  2009; 34: 195– 212. 39 Schnell R. Survey-Interviews: Methoden Standardisierter Befragungen (Survey Interviews: Methods of Standardised Assessments) . Wiesbaden: Springer VS, 2012. 40 Groves RM. Survey Errors and Survey Costs . New York: Wiley, 1989. Google Scholar CrossRef Search ADS   © The Author(s) 2018. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

The European Journal of Public HealthOxford University Press

Published: Apr 13, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off