Job stress and mental symptoms in forestry occupations

Abstract

Occupational stressors, job stress, mental job strain, and both mental and somatic symptoms in all the five occupations of forestry were clarified by inquiries in 1991 and 1994. The aim and the problems of this exploratory, comparative, and explanatory study were the following. Firstly, the comparison of the five occupational groups as regards stressors, certain coping resources, job stress, mental job strain, and mental and somatic symptoms. Secondly, the search for the main discriminators for these occupations. Third, an understanding-oriented evaluation of the potential effect of the structural change on stress-related problems. Fourth, the search for direct and indirect links between stressors, stress and outcomes, e.g. by the aid of correlations, regression analyses, and LISREL models. Fifth, search for moderators for stressor or stress-outcome relationships. Theoretically, psychological stress was defined as the perceived mental tension due to a cognitively appraised mismatch between the individual's internal or external resources or potentialities, on the one hand, and internal demands such as expectations, needs, values, goals or desires, or external demands set by the environment, on the other hand. Job stress was operationalized by asking, 'How stressed do you feel yourself at work', and mental job strain by asking, 'How mentally strained do you feel yourself at work'. The response rates and the numbers of respondents for the occupational groups studied without the additional short questionnaire for missing data were as follows: 80%, n = 556 in 1991, (74%, n = 244 in 1994) for Masters of Forestry, 85%, n = 1627, (78%, n = 258) for forest functionaries (engineers and technicians), 50%, n = 353, (52%, n = 173) for forest machine contractors, 58%, n = 172, (59%, n = 196) for forest machine operators, and 69%, n = 929, (68%, n = 224) for lumberjacks. The occupational groups were compared e.g. by the aid of analysis of variance, profile analysis and MANOV A. The main and interaction effects of the study timepoint and the occupational group on stress and symptoms were analyzed by the help of analyses of variance. Interaction effects of some moderator candidates on work pressure or job stress - mental symptoms relationship by occupation were clarified with MANOV A/ANCOV A techniques. The main discriminators for the occupations were determined with discriminant analyses. The explanatory models were based on LISREL and regression analyses. These regression models included also interactions of predictors and putative moderators. The structural changes in the forest sector have increased progressively from 1991 to 1994, but a downward trend is anticipated in the future. Fusions, reorganizations, and rapid mechanization has released not only manual labour, but also forest functionaries and Masters of Forestry. Discrepancies had increased systematically in every occupation, the Masters of Forestry scoring highest on this sum scale, followed closely by the functionaries. Along with the structural change, job security had weakened from 1991 to 1994 among other groups than machine contractors and operators. It is noteworthy that in 1994 the operators felt their job to be equally secure as the Masters of Forestry and more secure than their employers, i.e. the contractors. In 1991 Masters of Forestry were most exposed to role ambiguity, bullying, and poor organizational practice. These stressors were felt to be problems more often by white-collar than blue-collar workers. The functionaries, often acting between the forest workers and the Masters of Forestry, suffered most from job role conflicts. The weekly working time of contractors had increased by 6.5 hours and of operators more than 3.5 hours from 1991 to 1994. The difference in weekly working time between contractors and lumberjacks was more than 20 hours, and that between contractors and functionaries over 17 hours. In line with the elevated working time, also work pressure and the use of mental resources had risen crucially among contractors and operators, and slightly among Masters of Forestry during the three-year period. A lesser increase was found among lumberjacks, but no changes among functionaries. All in all, except for weekly working hours, job insecurity, and role conflicts, the Masters of Forestry scored highest on both process- and structure-oriented stressors. The operators scored lowest in many structure-oriented stressors, and the lumberjacks scored lowest in process-oriented stressors. As for process-oriented stressors (e.g. work pressure, weekly working hours, use of mental capacity) employees can recover from temporary quantitative job overload by resting, and process-oriented stressors do not necessarily affect mental well-being outside work. Structure-oriented stressors (e.g. discrepancies, structural change, job insecurity, bullying, role conflicts, role ambiguity), however, often extend their adverse effects also onto leisure time and the entire well-being. As regards resources, no considerable occupational differences were noted in sense of coherence, hostility, or self-esteem in 1991 nor in 1994. White-collar groups scored highest in perfectionism. Type A behaviour pattern increased on a continuum from blue-collar to white-collar occupations. In 1991 Masters of Forestry scored highest on job supplies (good properties of job) followed by functionaries, but in 1994 contractors perceived their job supply as better than the functionaries, and also operators and lumberjacks came close to the functionaries. From 1991 to 1994 white-collar employees had lost their job supplies, but blue-collar groups had moved in the opposite direction. In both study years, Masters of Forestry felt most job stress and mental job strain and lumberjacks the least. All in all, in spite of the clearly elevated job stress from 1991 to 1994, mental job strain had not increased. However, in line with the rise in job stress, there was also a modest rise in mental symptoms from 1991 to 1994. The mental symptoms of lumberjacks had increased most, and they scored roughly at the level with the contractors and Masters of Forestry in 1994. While the job stress and clearly also the mental job strain of operators had increased, their symptoms had slightly diminished from 1991 to 1994. No clear change was found in the mental or somatic symptoms of the functionaries. Their situation has remained stable from 1991 to 1994 as regards most stressors, moderators, job stress, mental job strain, and symptoms. Otherwise, the rest of the groups had more somatic symptoms, but operators only marginally. The greatest increase in somatic symptoms was found in contractors followed by lumberjacks. The Masters of Forestry, although scoring highest in mental symptoms, scored lowest in somatic symptoms. Stepwise discriminant analyses revealed that process-oriented stressors exerted a stronger discriminative power than structure-oriented stressors. Job supplies and Type A behaviour pattern also proved to be powerful discriminators for the occupations. As a variable block, outcome variables (job stress, mental job strain, mental and somatic symptoms) had fairly modest, though most often significant, discriminative power compared to stressors and certain moderators. Summing up, outstanding occupational differences in several stressors and certain moderators weakened or turned into, distinctions without differences, when dealt with outcome variables. The reasons for this finding might be e.g. that resources or buffers may moderate stressor-outcome relationships. When moving towards white-collar occupations, also several resources showed an upward trend. The occupational differences in outcomes might also be obscured by the sleeper effect, the accumulation effect, the adjustment effect, or the healthy worker effect, or regression to the mean. Although statistically significant interoccupational differences arose also in the outcomes, these differences might psychologically be of minor importance. The regression models for job stress, mental job strain, and for their sum displayed a high degree of similarity. When entered after inclusion of stressors, potential moderators explained only 3.2-4.4% of the variance of these dependent variables, and interactions explained 2.2-2.4% of the remaining variance. Mental job demands and work pressure were the most powerful predictors of mental job stress. Mental, somatic and all the symptoms were best predicted by potential moderators, especially by personal characteristics. If entered after the moderator block, stressors raised the explanation grade by 5.7-6.6%. The most powerful predictors for mental symptoms in 1991 were coherence, hostility, self-esteem and job supplies from moderators and, and on the other hand, discrepancies, mental job demands, bullying and work pressure from stressors. The explanatory power of work pressure was prominent for job stress, but not for symptoms. The contribution of interactions varied around one percent of the variance, and the interactions showed no clear tendency. As for the interactive effects of potential moderators on work pressure or job stress - symptoms relationship, occupational differences arose, but they followed no logical pattern. The results of the interactions seemed to be sensitive to the statistical procedures used. The explanation degrees varied around 40-50%. The links between stressors, stress and symptoms were ascertained by LISREL models. The most prominent distinction between these two kinds of analyses was that discrepancies did not have a stable and strong link with mental symptoms, as suggested by regression analyses. In forestry occupations job stress and mental symptoms were evidently not more prevalent than in other occupations. The rise in mental symptoms was in line with many earlier studies concerning both the forest sector and other fields of economic activity.