Abstract
BACKGROUND:
Shift working is unavoidable in many industries with continual material processing such as petrochemical plants. So, the adverse effects of shift working on workers should be seriously considered.
OBJECTIVE:
This cross-sectional study evaluated occupational fatigue and mental health complaints and their relationship in rotating 8-hour shift workers.
METHOD:
In this study, 287 shift workers participated. The Multidimensional Fatigue Inventory (MFI-20) and General Health Questionnaire (GHQ-28) were used to evaluate the level of fatigue and mental health, respectively.
RESULT:
A relatively high prevalence of mental health complaints (particularly social dysfunction) and fatigue (especially general fatigue) were found among the study population. In general, 43.4% of participants reported a mental health problem. A moderate correlation was found between fatigue and mental health (r = 0.58). The stepwise regression model revealed that fatigue was significantly related only to “anxiety and insomnia” and “severe depression”.
CONCLUSION:
This study revealed that the 8 h shift workers in studied areas are exposed to a considerable risk of mental health and fatigue. So, improving the ergonomics and health aspects of the workplace is recommended to reduce related risk factors.
Introduction
The term “shift worker” is generally used for individuals who work any schedule other than the standard daylight hours (7 am–6 pm) [1]. Fifteen to thirty percent of the working population in industrialized countries work as shift workers [2], while this rate seems to be higher in industrially developing countries due to some issues such as poor job design and organization, inappropriate working schedules and night working [2, 3]. Evidence suggests that day/night shift rotation schedule is an important aspect of shift working which can result in circadian rhythm and sleep disorders [4, 5]. In addition to cardiovascular and gastrointestinal diseases, shift working is one of the main contributors to fatigue [6, 7]. Prolonged sleep deprivation and cumulative exhaustion can cause chronic sleeplessness in shift workers. Consequently, this may lead to permanent sleepiness, decreased alertness and efficiency, and increased fatigue, failure and incident rate among shift workers compared to other workers. This situation not only influences the workers’ health and satisfaction, but also burdens the families and society with emotional, social and economic problems [8–11].
Fatigue is generally recognized by several symptoms such as weakness, energy defect, burnout and decreased physical and cognitive activity [12]. Fatigue is also one of the most important symptoms of distress, which is a common experience caused by insufficient rest or sleep, heavy physical or mental activity, and lack of motivation to begin any activity [13].
Several studies have demonstrated that fatigue is a usual complaint. According to findings of a cross sectional study conducted among the working population of fifteen European countries, 5% to 56% of workers suffered from fatigue [14]. The impact of work-related fatigue on the performance, safety and work ability has been evaluated in some studies [15, 16].
Mental health is another important issue that can be considered a consequence of shift working. Several studies have investigated fatigue and mental health in shift workers [17, 18]. Åhsberg (2000) reported a higher prevalence of fatigue among night shift workers compared to other groups [19]. The authors also found that an increase in reaction time was related to the mental aspects of fatigue. Similar studies among various working populations such as anesthesiologists have reported a significant correlation between mental health and fatigue [20, 21]. Notwithstanding, other studies have reported no relationship between mental health and shift working [22, 23]. So, it seems that more investigations are needed in this respect.
In the majority of industries with continual production processes such as petrochemical industries, shift working is necessary. Consequently, the risk of adverse health effects from shift working should be seriously considered [4, 24]. According to our findings, no study has been reported about fatigue and its relationship with mental health in petrochemical shift workers. Therefore, this study aimed to evaluate fatigue and mental health and their relationship in 8-hour shift workers of petrochemical industries.
Methods
Design and sample
This cross-sectional study was carried out between September 2013 and April 2014, in a petrochemical company in Mahshahr, Iran. The study population was 396 male workers who worked in rotating 8-hour shifts. Their mean (±SE) age and shift work experience were 34.27±0.4 years and 8.91±0.39 years, respectively. The participants with serious physical or mental illness (according to their responses to a question in questionnaire and also medical recordings) were excluded from data analysis. The shift working system consisted of three 8-hour shifts (morning: 6:00 am to 2:00 pm; afternoon: 2:00 pm to 10:00 pm; and night shift: 10:00 pm to 6:00 am). Also, the shift schedule was in a 3M–3A–3N–3O pattern (3 mornings–3 afternoons–3 nights–3 off) in a forward rotating program. A questionnaire consisting of three sections was used for data collection, covering the following items: demographic characteristics (including age, body weight and height, number of children, marital status, educational level, shift work tenure and job title); multidimensional fatigue inventory (MFI) and general health questionnaire (GHQ). The questionnaire was distributed among all shift workers in all occupational groups including Health, Safety and Environment (HSE), operation, maintenance, engineering, office work and security departments. Due to limited shift working population in the studied site, all of the shift workers were included in the study. The study participants were familiarized with the purposes of the study and instructed in how to complete the questionnaire. The completed questionnaires were collected in collaboration with the foremen. A total of 287 participants voluntarily completed the questionnaires, resulting in a response rate of 72.5%. Each participant completed a consent form before participating in the study. The ethical review committee of the Tabriz University of Medical Science reviewed and approved the study protocol.
Measurement
Fatigue
The Multidimensional Fatigue Inventory (MFI-20) is a 20-item self-report assessment tool to measure fatigue [25]. It consists of five dimensions of fatigue: general fatigue, physical fatigue, reduced motivation, mental fatigue and reduced activity. A 5-point Likert scale is used for each item (scores ranging from 4 to 20 for each dimension), so that the total score of fatigue ranges from 20 to 100. Each dimension consists of 4 items, two indicative for fatigue and two contraindicative. For the indicative questions, a high value reveals a high fatigue level and for the contraindicative questions, a high score shows low fatigue level. “General fatigue” refers to the overall functioning status which is expressed by the statement: “I tire easily”. This dimension is able to evaluate both aspects of physical and psychological fatigue. “Physical fatigue” covers the physical feeling of fatigue. The “mental fatigue” scale measures cognitive function, such as having trouble in concentrating. “Reduced activity” measures the effects of both mental and physical risk factors on the activity rate. “Reduced motivation” refers to feelings that may prevent an individual from starting any activity [25, 26]. This self-report tool has an established reliability and validity and is one of the most frequently used questionnaires to measure fatigue [25–27]. In this study, the English version of MFI–20 has been translated and revised into Farsi and has an established validity and reliability [28, 29].
Mental health
The General Health Questionnaire (GHQ-28) [30] is widely used as a self-report assessment tool to evaluate mental health. The GHQ-28 consists of four subscales: somatic symptoms (items 1 to 7), anxiety and insomnia (items 8 to 14), social dysfunction (items 15 to 21) and severe depression (items 22 to 28). Each item is scored based on a 4-point scale ranging from “never true” (score of 0) to “always true” (score of 3), and scoring from 0 to 3. The scores varied from 0 to 21 for each subscale and from 0 to 84 for the total GHQ-28 (which was defined as mental health in this study). Higher scores on this scale show the presence of more severe mental problems [30]. In this study, the Persian version of the GHQ-28 was used. With a cut-off score of 23, the sensitivity, specificity and overall misclassification rate for a GHQ–28 were 70.5% ±2.7, 92.3% ±2.4 and 12.3% ±2.4, respectively [31]. Due to transformation of score ranges as described in the statistical methods, the cut-off score of GHQ-28 was changed to 27.38.
Statistical methods
Data were analyzed using SPSS version 11.5 [SPSS Inc., Chicago, IL, USA]. To compare the GHQ and MFI results, the score range of all subscales in GHQ-28 and MFI-20 were transformed into 0 to 100. The internal consistency of questionnaires was evaluated with Cronbach’s alpha. Based on the average inter-item correlation, the normality of data was tested by Q-Q test for all continuous variables. All descriptive values were shown as Mean±SD at each time interval. Stepwise regression, R square and Pearson correlation test were performed to assess the relationship between fatigue and mental health. For better correlation analysis, the scores were categorized in “low”, “medium” and “high”. In addition, error bars were used to show confidence intervals, differences and relationships between the mean scores. The level of significance was set at P-value < 0.05.
Results
Demographic data
A total of 287 workers filled out the questionnaires. The demographic characteristics of the study population are shown in Table 1. As the data show, the operation workers were the group with the highest frequency distribution. The majority of participants were married and more than half of the study population (58.5%) had an academic education level. The data also showed that the shift work experience among workers (8.91±0.39 years) was adequate to investigate their occupational health status.
Demographic data of the study participants (n = 287)
Demographic data of the study participants (n = 287)
Table 2 lists the subscales scores and Pearson’s correlation of the MFI-20 among the studied workers. The total mean fatigue score of the study population was 36.07 (SD = 16.45). General fatigue, with a mean score of 52.54 (SD = 21.2), and reduced motivation, with a mean score of 26.06 (SD = 18.14), had the highest and lowest mean values, respectively. As can be seen from Table 2, there were moderate positive correlations between the MFI subscales, with r values ranging from 0.32 to 0.66. There was a strong positive correlation between reduced activity and reduced motivation (r = 0.66), while in contrast there was a weak positive correlation between general fatigue and reduced activity (r = 0.32). The correlation coefficients between the MFI total score and its subscales varied from 0.74 to 0.85. The Cronbach’s α values were obtained for the MFI subscales as follows: general fatigue: 0.68; physical fatigue: 0.62; reduced activity: 0.73; reduced motivation: 0.57 and mental fatigue: 0.70. The total MFI with 20 items had a Cronbach’s α value equal to 0.88, which indicates a good internal consistency for the MFI-20 and its subscales.
Mean and standard deviation and Pearson’s correlation between the subscales of the MFI-20 (n = 287)
Mean and standard deviation and Pearson’s correlation between the subscales of the MFI-20 (n = 287)
The scores of the GHQ-28 and its subscales are shown in Tables 3 and 4. It can be seen from these tables that the value of 27.38 was considered as the cut-off point for GHQ and its subscales. It was revealed that 43.4% of the study population reported some kind of mental health issues, with a mean score of 29.49 (SD = 16.28) (range = 0–100). Social dysfunction (92.4%), anxiety and insomnia (52.4%) and somatic symptoms (47.6) had the highest prevalence among workers with the mean scores of 39.34 (SD = 10.85), 33.22 (SD = 23.06.36) and 29.21 (SD = 21.03), respectively. In contrast, only 18.6% of workers complained of severe depression, with a mean score of 15.34 (SD = 21.53). A Pearson’s correlation test was performed to evaluate the inter-relationship between GHQ subscales. As shown in Table 3, there was a moderate to high positive correlation between the GHQ subscales (r values ranging from 0.51 to 0.79). The correlation coefficients between the GHQ total score and its subscales ranged from 0.82 to 0.88. The Cronbach’s α values was obtained for the GHQ subscales as follows: somatic symptoms: 0.88; anxiety and insomnia: 0.90; social dysfunction: 0.71; severe depression: 0.92. The total GHQ (which was considered a mental health indicator) with 28 items had a Cronbach’s α value of 0.94. These values show a good internal consistency in the GHQ-28 and its subscales.
Score on the General Health Questionnaire (GHQ-28) (n = 287)
Score on the General Health Questionnaire (GHQ-28) (n = 287)
Mean and standard deviation of the GHQ-28 and Pearson’s correlation between the subscales of the GHQ-28 and total MFI (n = 287)
A stepwise regression test was used to evaluate the relationship between mental health (GHQ total score) and fatigue (MFI subscales scores). As can be seen from Table 5, only general fatigue, physical fatigue and mental fatigue remained in the model (P values < 0.05), while reduced motivation with reduced activity were excluded. Thus, the final model is as follows:
Also, a stepwise regression test was used to evaluate the relationship between total fatigue (MFI-20 total score) and GHQ-28 subscales. As shown in Table 6, only anxiety and insomnia, and severe depression remained in the model (P values < 0.05) and somatic symptoms with social dysfunction were excluded. Thus, the final model is as follows:
In addition, to diagnose the relationship between GHQ and MFI, a linear regression test was performed with R square = 0.583 and Adjusted R square = 0.337, score. Thus, the final model is as follows:
This means that per each unit score of MFI, 0.582 is added to the GHQ score.
Regression relationship between the mental health problem (GHQ total score) and fatigue (MFI subscales scores)
Regression relationship between the mental health problem (GHQ total score) and fatigue (MFI subscales scores)
R square = 0.624, Adjusted R Square = 0.383.
Regression relationship between the total fatigue (Total MFI-20) and subscales of mental health
R square = 0.593, Adjusted R Square = 0.348.
As shown in Table 4, the Pearson’s correlation test between the GHQ subscales and total fatigue indicate a fairly positive correlation between total fatigue with anxiety and insomnia (r = 0.58) and somatic symptoms (r = 0.53). However, a weak positive correlation was found between total fatigue and social dysfunction (r = 0.37). The relationship between total fatigue and mental health was also evaluated with the Pearson’s correlation test, which showed a relatively high positive correlation (r = 0.58).
Figure 1 presents the relationship between the total MFI and GHQ subscales scores. The GHQ subscales scores in workers with moderate to high level of total fatigue (3–5 on the 5-point scale) were significantly higher than those with low level of total fatigue (1-2 on the 5-point scale).
Relationship between the total fatigue score of the MFI and GHQ subscale score.
Similarly, the MFI-20 subscales scores were significantly higher in those workers with the GHQ score≥27.38 than those with a lower score (GHQ score≤27.38) (Fig. 2).
Relationship between levels of fatigue dimensions with total GHQ-28.
The main purposes of this study were to determine the prevalence of mental health complaints and fatigue, and to evaluate their relationship among petrochemical 8 h shift workers. The GHQ results showed a relatively high prevalence of mental health problems (43.4%), particularly in the form of social dysfunction (92.4%) and anxiety and insomnia (52.4%) among the studied 8 h shift workers. These prevalence rates are in agreement with values reported by other studies among working populations, 84% in anesthesiologists [21], and 47.5% in a Polish working population [32]. Another study, previously conducted by the authors among 12-hour petrochemical shift workers of the southern Pars gas field in Iran, showed higher psychological distress (total GHQ mean score, 34.66) [33]. The poor working and bad weather conditions, prolonged shift hours, lack of leisure facilities and being away from family in the Pars gas field are the most probable factors in this area. It seems that the prevalence values of distress can vary depending on the study population, method and tools. The increased social dysfunction rate in 8 h shift workers reported in our study (mean score of 39.34) is not in agreement with the prevalence rate of 14.2% reported in the general Iranian population [34]. The prevalence of mental health problems of the studied population was 43.4%, which is considerably higher than the value reported in the general Iranian population (20.9%) [34]. This may be associated with occupational health and safety risk factors in petrochemical industries such as high work load, chemical hazardous and explosive materials, harmful noise exposure, stressful working conditions, and insufficient off-the-job social interactions. Also, an epidemiological study on the Iranian population reported a prevalence rate of 8.48% of anxiety [35], while in the present study this rate is dramatically different. A significant association has also been reported between shift working and anxiety among nurses [36]. The relatively high positive correlation between the GHQ subscales (r values ranging from 0.51 to 0.79) shows a high common variance proportion inherent to all GHQ subscales and their good relations.
The MFI results indicated a high general fatigue, moderate physical and mental fatigue, and low reduced motivation among 8 h shift workers, which is consistent with studies conducted among various working groups such as anesthesiologists [21]. The results of a previous study conducted by the authors in petrochemical 12-hour shift workers of the southern Pars gas field in Iran indicated relatively higher fatigue levels (total MFI-20 mean score, 42.6) [33]. As mentioned earlier, this can be related to some factors such as poor working conditions, prolonged working hours and etc. Unfortunately, there was no other study on occupational fatigue in shift workers via MFI-20 with which to compare our results. However, in comparison with general population studies, the fatigue rates reported by workers in the current study were considerably low. For example, the mean score of 52.54 was obtained for the general fatigue in our study, while the mean value of 35 have been reported in Danish general population [37]. On the other hand, compared to a study conducted on cancer patients, the mean scores of all fatigue dimensions in our study population were dramatically low [38]. The comparison of these findings reveals that the total fatigue in our working population is higher than the general population and lower than patient groups. Since there are no certain defined cut-off values for the MFI score in the literature, it was not possible to provide and report the prevalence values of fatigue among other groups.
In this study, a relatively high correlation was found between the MFI-20 subscales (with r values ranging from 0.32 to 0.66), which are somewhat comparable with values (0.49 to 0.74) reported among American adults [39]. However, different correlation coefficients (0.012–0.48) were reported for fatigue dimensions in the study on a cancer patient group [38]. In our study, the highest correlation was obtained between reduced activity and motivation (0.66), while the highest correlation in the study on cancer patients was reported between general fatigue and physical fatigue. Perhaps, the treatment program difficulties such as long-term hospitalization resulted in increased physical stresses among patients, which may chronically influence mental status. On the other hand, shift work arrangement and its patterns in the Iranian petrochemical sites, particularly in the studied region, may have had a significant impact on the reported high score of mental health in this study. In addition, our study revealed that there is a strong positive correlation between physical and mental fatigue in shift workers, while a weak correlation was reported between these subscales in the study by Munch (2006) among cancer patients [40]. It should be noted that the differences in the fatigue values could have been affected by the differences in gender, age or educational level of the respondents, and thus further studies are needed to clarify the etiology of fatigue.
The findings of this study showed a relatively high correlation between the total GHQ-28 and total MFI-20 (r = 0.58). This means that the shift workers who reported higher levels of mental health problems experienced higher levels of fatigue (Fig. 1). This finding is in agreement with the results of the study on anesthesiologists [21]. In a population based study in England, the GHQ-12 and a fatigue questionnaire were used to evaluate mental health and fatigue [20], and their findings indicated a fairly high positive correlation (r = 0.62), which is similar to our results. The stepwise regression showed that mental health was associated only with general fatigue, physical fatigue and mental fatigue. This means that reduced motivation and reduced activity had no significant effect on mental health in 8 h shift workers. Also, the stepwise regression model revealed that total fatigue was significantly related only to “anxiety and insomnia” and severe depression. Increased mental health problems in workers with moderate to high levels of total fatigue (Fig. 1) and an increased rate of all fatigue dimensions in the group with a GHQ score above the cut-off point showed a two-way relation between mental health status and fatigue (Fig. 2). It is interesting to note that despite relatively high total fatigue and staying away from family, the studied workers reported very low depression and reduced motivation, which may be attributed to higher income and promising religious beliefs. Based on the results, the 8 h shift workers in the studied region are exposed to a considerable risk of severe mental health. Previous studies have suggested the necessity of corrective measures and recovery from fatigue to improve shift work systems [41]. Shift scheduling together with other effective variables such as occupational hazards, environmental conditions and psychosocial factors can be considered as determinants. Other studies in this area are needed to evaluate the effects of these parameters.
There are several limitations that should be taken into account before generalizing the results of this study to other shift working populations. This study was conducted in a workplace with special environmental conditions (e.g. high temperature and relative humidity levels), and this may have affected the fatigue and health status of the workers. A future case-control study may help to compare results for understanding what differences between groups there may be for the studied problems. In addition, since the MFI has been used generally in numerous patient group studies, it was difficult to compare our results with other studies. Thus, further studies in other working environments and settings using the same instruments are recommended. Finally, a subjective method was used to collect the required data, and subjective measures may be susceptible to individual motivation.
Conclusion
In conclusion, the present study showed a relatively high prevalence of mental health concerns (particularly social dysfunction) and fatigue (especially general fatigue) in 8 h rotating shift workers at a petrochemical plant in Iran. There was a fairly high positive correlation between mental health and fatigue in the study population. The prevalence rates of mental health and fatigue and their relationship in the present study were found to be different from those reported among general population and patient groups. These findings highlight the need for ergonomic interventions to improve the working conditions of shift workers in similar working environments.
Conflict of interest
All authors have no conflicts of interest to declare.
Funding
No fund was received for this study.
Footnotes
Acknowledgments
This manuscript was extracted from the thesis written by Ahmad Bazazan, MSc of ergonomics. The authors wish to acknowledge the support provided by department of health, safety and the environment (HSE) of Razi Petrochemical Company and all shift workers participated in this study. We thank Richard Gibbons, Ergonomist in Washington State Department of Labor and Industries, for comments that greatly improved the manuscript.