Determinants of Sense of Coherence in Male Japanese Workers : a Cross-Sectional Study

Background: Suicide is the leading cause of death of Japanese people aged 20-39 years. Improved stress-coping ability, here defined as sense of coherence (SOC), may reduce suicidal risk. The objective of the present study is to investigate the determinants of SOC, including age, in male Japanese workers. Methods: Of the 467 employees of a nationwide company in Japan, 457 agreed to participate in this cross-sectional study. After exclusion of 43 females, the remaining 414 males were analyzed. SOC was measured by the SOC-13, and other data were collected using a self-administered questionnaire. Multiple linear regression analyses were performed to evaluate possible determinants of SOC. Results: Higher age was significantly associated with increased SOC (β=0.14, p=0.014). Workers in their 40s and 50s had 2-3 points higher adjusted mean SOC scores than did workers aged 20-39 years. However, the most influential factor for SOC was sleep quality. Marital cohabitation status (living with a spouse) and drinking habits (current drinker) were also associated with higher SOC. In contrast, smoking habits, exercise habits, and night shift work were not significantly associated with SOC. Conclusions: Our findings suggest that age, quality of sleep, drinking habits and marital cohabitation status may be mediators of SOC.


Introduction
In recent decades in Japan, suicide rates have increased among people in their 20s and 30s, but have decreased in other age groups.Suicide is now the leading cause of death in Japanese people aged 20-39 years (http://www8.cao.go.jp/jisatsutaisaku/ whitepaper/w-2014/html/index.html).In addition, psychological stress is associated with increased risk of suicide 1) .Because young people often have not acquired good strategies for dealing with stress, improved stress-coping ability may reduce suicidal risk.
Sense of coherence (SOC) was proposed by medical sociologist Aaron Antonovsky as the ability to cope with stress or crises and derive something positive from them 2) 3) .SOC consists of three faculties: comprehensibility, manageability and meaningfulness, and is measured by a self-administered questionnaire.These faculties are the cognitive, instrumental and motivational components of SOC, respectively 4) .The combined score for the three components is calculated as the SOC score, where higher values indicate stronger SOC.SOC is not an inborn or innate capacity, but a learned sense 5) .Research has shown that strong SOC is associated with various factors such as older age 3) 6)-8) , male sex 6) 8) , high income 8) and good health 8) -11) .
Although several predictors of SOC have been studied in other countries, results are not always in agreement.In Japan, many SOC studies have used female nursing students as their subjects; thus, the determinants of SOC in Japanese men remain to be elucidated.A previous study in male Japanese workers 7) reported that higher age, quality of sleep, eating vegetables every day and working shorter hours were important for enhancing SOC, although the analysis did not always fully control potential confounders.Accordingly, a better understanding of lifestyle factors that improve SOC is an urgent issue, particularly in countries with high suicide rates, such as Japan.Therefore, we decided to study the relationship among SOC, age and a variety of lifestyle factors in Japanese men, with an objective of investigating the determinants of SOC in male Japanese workers.

Participants
A cross-sectional questionnaire was distributed among workers at a Japanese nationwide petroleum-related company in 2013.Of 467 employees, 457 agreed to participate (participation rate = 97.9%).Most participants were white-collar workers.Written informed consent was obtained from all participants when they entered the study.There were 414 males and 43 females.Only the male sample was used for analysis because the female sample was too small to analyze.The mean age of males was 42 years and the median was 44 (range, 20-59) years.The study procedure was approved in advance by the Institutional Review Board of the Juntendo University Faculty of Medicine, Tokyo, Japan (receipt number 813; approval letter number 2012057; May 21, 2012).

Data collection
The questionnaire comprised 39 questions including sociodemographic and anthropometric characteristics and lifestyle and exercise habits in addition to Japanese versions of the SOC 13-item scale 2) and the Pittsburgh Sleep Quality Index (PSQI) 12) .Higher values of SOC score indicate stronger SOC.A PSQI value > 5 indicates"poor sleeper" .Participating in regular exercise was defined as exercising at least twice a week for 30 minutes per session, for more than 1 year.This definition was also employed in the National Health and Nutrition Survey in Japan by the Ministry of Health, Labour and Welfare 13) .Because night shift workers may have higher SOC to cope with stress, night shift work was also investigated.

Statistical analysis
The authors constructed two types of age variables: one continuous and the other categorical.To create the categorical variable, we divided participants into three age groups: 20-39, 40-49 and 50-59 years.Medians of each group were 29, 45 and 55 years, respectively.For drinking and smoking habits, dummy variables defined current smokers or drinkers as a reference category.For the quality of sleep, we calculated global PSQI score.Adjusted means of SOC and its three components by age group were calculated using the SAS GLM procedure.These analyses were adjusted for exercise habits, drinking habits, smoking habits, sleep quality, marital cohabitation status (married and living with spouse, or not) and night shift work.The authors also conducted multiple linear regression analyses using age (continuous), current  exercise habits, drinking habits, smoking habits, quality of sleep, marital cohabitation status (married and living with spouse, or not) and night shift work as predictor variables and SOC score as the dependent variable.The highest variance inflation factor was 1.3.Missing data were handled by complete case analysis.In the end, 352 participantsʼ data were entered into the regression analysis.All statistical analyses were two-sided and significance was accepted at p < 0.05.SAS Enterprise Guide 6.1 (SAS Institute, Cary, NC, USA) was used for the statistical calculations.

Participant characteristics
Table-1 shows the characteristics of the 414 male workers.Most workers were current drinkers and 39.4% were current smokers, while 57.1% were married and living with their wives.Most workers worked the day shift.Mean body mass index and PSQI values were normal and somewhat poor, respectively.There was high internal consistency for the SOCʼs 13 items (Cronbachʼs α = 0.829).

Age and SOC
Table-2 presents adjusted mean SOC scores by age group.Scores for workers aged 20-39 years were about 2-3 points lower than those of the other two groups.Regarding the three elements of SOC, comprehensibility had the biggest difference, i. e., increased comprehensibility was associated with higher age.

Other factors and SOC
Table-3 shows the results of multiple regression analysis using SOC as a dependent variable.Higher age was significantly associated with increased SOC (β = 0.14, p = 0.014); however, the most influential factor in SOC was sleep quality, according to the standardized partial regression coefficient.Marital   status (living with a spouse) and drinking habits (current drinker) were also associated with higher SOC.In contrast, smoking habits, exercise habits, and night shift work were not significantly associated with SOC.Similar results were obtained even when age was categorical (20-39, 40-49, and 50-59 years) rather than continuous (data not shown).
Although the frequency of current night shift work [nights/month] was inversely correlated with SOC score (Spearman r =-0.12, p = 0.016), this association was attenuated to null after adjustment for covariates in multivariable regression analysis (data not shown).

Discussion
Higher age was associated with elevated SOC in Japanese men.In particular, increased comprehensibility was found in higher age groups.An accumulation of life experience with age may strengthen the ability to view life as comprehensible.The strongest influence on SOC was sleep quality, while living with a spouse was also associated with higher SOC.These factors may be effective for developing SOC.In other words, good sleep and cohabitation may promote psychological resilience.Past drinkers had an 11.5-point lower SOC than current drinkers; this difference may reflect the existence of a life event, such as disease, that forced them to stop drinking.Never drinkers had 3.7-point lower SOC than did current drinkers, and this may signify the phenomenon that most people can obtain some relief from their stress by drinking although there is no available literature related to this topic as far as we know.However, we do not recommend alcohol drinking to increase SOC as many epidemiologic evidences suggest that current heavy drinking habit is not good for health.The lack of significant associations of SOC with smoking habits, exercise habits, or night shift work suggests that these factors may not play a critical role in development of SOC.In particular, smoking will not improve SOC.To the best of our knowledge, this is the first study to examine the association between night shift work and SOC.This investigation is important because night shift may cause sleep disturbance and act as stressor.
The positive association between age and SOC in the present study is consistent with some previous cross-sectional studies in Japan 7) and Sweden 6) as well as a systematic review that showed a tendency for SOC to increase with age over the whole life span 3) .Although some follow-up studies over a few years have shown small differences in mean SOC over time, they did not observe SOC over a longer period 6) .As in the present study, SOC increased with increasing age and with sleep quality in another male Japanese cross-sectional sample 7) .With regard to cohabiting with a spouse, our findings agree with those of a previous study, which showed that people sharing households with one or more persons report more favorable SOC scores than do those living alone 8) .The lack of significant association of SOC with smoking is consistent with results from a Finnish student study 14) , but inconsistent with a Norfolk (UK) study that showed an inverse relationship between these variables 15) .Association between alcohol drinking and SOC was also observed in the Norfolk study 15) , whereas SOC had no relation to drinking among students in a Finnish study 14) .The lack of significant association between exercise habits and SOC is also inconsistent with previous studies 14) 15) that have shown positive associations between physical activity and SOC.The reason for such marked inter-study differences is still unclear, but they may be attributable in part to the differences in social background or controlled covariates.Further studies are needed to confirm replication of these associations.
According to our additional analysis using each element of SOC as a dependent variable, the same factors mentioned above were associated with comprehensibility.On the other hand, age was not significantly associated with manageability.Participation in regular exercise and having no experience of working night shift were also associated with higher meaningfulness (data not shown).Such a subcomponent analysis can also be informative.
Major strengths of the present study include high participation rate and the control of confounding factors.First, almost all invited subjects participated in the study, likely eliminating the possibility of nonresponse bias.Second, many potential confounders were measured and statistically controlled by multivariable analysis.Mean SOC levels in different age groups were also adjusted for covariates.
Several limitations also warrant mention.First, as this is a cross-sectional study, we cannot necessarily infer whether the aforementioned factors caused the changes of SOC.Second, we cannot rule out the possibility of measurement error due to the use of self-reported SOC, which may have influenced the lack of observed associations in several cases.Third, we did not measure vegetable intake despite its possible association with the"meaningfulness"component of SOC observed for Japanese male workers in a previous study 7) .This segment of the population will require investigation in future research.

Conclusions
SOC increased with age in males, and was also associated with quality of sleep, drinking habits and marital cohabitation status.Our findings suggest that these factors may mediate development of SOC.Accordingly, improving sleep quality, living with a marital partner and drinking moderately might be prudent recommendations for increasing SOC, particularly for young men.Moreover, the message that smoking is not necessary for improving SOC is one worth sharing.

Table - 1
Characteristics of study participants (414 male workers)

Table - 2
Sense of coherence (SOC) and its three components in different age groups

Table - 3
Results of multiple linear regression analysis (influences of each predictor variable on sense of coherence [SOC] score)