Excessive Daytime Sleepiness and its Associated Factors among Male Non-shift White-collar Workers

Transcription

1 J Occup Health 2002; 44: Journal of Occupational Health Excessive Daytime Sleepiness and its Associated Factors among Male Non-shift White-collar Workers Yuriko DOI, Masumi MINOWA and Toshiharu FUJITA Department of Epidemiology, National Institute of Public Health, Tokyo, Japan Abstract: Excessive Daytime Sleepiness and its Associated Factors among Male Non-shift Whitecollar Workers: Yuriko DOI, et al. Department of Epidemiology, National Institute of Public Health Excessive daytime sleepiness (EDS) has been noted as a tremendous burden on our modern society and life, but evidence on EDS is limited for white-collar workers in occupational settings. To estimate the prevalence rate of EDS and examine its associated factors, we investigated 3,909 male non-shift whitecollar employees aged working in a telecommunications company in the Tokyo metropolitan area between December 1999 and January 2000 (a response rate of 91.0%). Our main findings in the present study were that the prevalence rate of EDS was 7.2% and five factors associated with EDS were identified in a multivariate logistic regression model (p<0.05): nocturnal sleep duration, sleep-wake schedules, depression, marital status and smoking. These results suggest that the individual behavioral modification of sleep habits and lifestyle is one of the promising strategies for reducing EDS. More importantly, it is possible that occupational and health promotion policies such as a ban on overtime work and the provision of mental health hygiene and social support at worksites are effective for the prevention of EDS in the workplace. (J Occup Health 2002; 44: ) Key words: Excessive daytime sleepiness (EDS), Epworth Sleepiness Scale (ESS), Prevalence, Associated factor, White-collar worker, Sleep duration, Sleep-wake schedule, Depression, Marital status, Smoking The consequences of sleep deprivation and sleepiness have been noted as a tremendous burden on our modern society and life 1), including an increase in mortality and Received Oct ; Accepted Jan 31, 2002 Correspondence to: Y. Doi, Department of Epidemiology, National Institute of Public Health, Minami, Wako, Saitama Japan morbidity, errors and accidents, absenteeism, a decrease in productivity, and the deterioration of personal and professional relationships 2, 3). Japanese society is not an exception with respect to sleep loss and sleepiness 4). In particular, excessive daytime sleepiness (EDS) is the critical issue in workplaces from the viewpoints of occupational health and safety. Past research has focused on daytime sleepiness and alertness among shift workers whose circadian rhythms are affected by the irregular sleep-wake schedules due to their shift work 5, 6), but evidence of EDS among non-shift white-collar workers is very limited though most of the workers are not engaged in shift work. Therefore, the purpose of our study was to shed light on EDS among male non-shift white-collar workers residing and working in the Tokyo metropolitan area, who restrict their sleep due to work, work-related activities and their long commute 7, 8). The specific aims of this study were to estimate the prevalence rate of EDS and examine the factors associated with EDS. As for the putative factors associated with EDS, sociodemographic factors (e.g. gender, age, educational status, type of work and marital status), sleep factors (e.g. sleep deprivation, sleep fragmentation and sleep-wake schedules), depression, physical state and psychoactive substances (e.g. medication to help sleep, alcohol, caffeine and cigarettes) were considered 5, 9). Subjects and Methods 1. Subjects The telecommunications company employed 5,608 white-collar workers at 22 work sites distributed in Kanagawa Prefecture at the time of the survey conducted between December 1999 and January After permission for the study was obtained from the headquarters and the labor union, questionnaires on sleep and health were distributed to 5,571 full-time white-collar workers by site staff. A total of 5,072 workers responded to the survey (a response rate of 91.0%). Each questionnaire, sealed in an envelope, was collected from

2 146 J Occup Health, Vol. 44, 2002 a respondent by site staff and directly sent to the authors. This data collection procedure kept the information obtained from the survey from other individuals in the company. The subjects analyzed were 3,909 male nonshift white-collar workers (Table 1). Female workers and workers on rotating shifts or permanent night workers were excluded from the analyses, since their lifestyle in the worksites and at home was different from male nonshift white-collar workers. Those who had any missing responses to the items on gender and age were also eliminated from the analyses. 2. Measurements Excessive daytime sleepiness was assessed by means of the ESS 10). The ESS is a self-administered questionnaire to measure sleep propensities in eight different situations of real-life: 1) sitting and reading, 2) watching TV, 3) sitting, inactive in a public place (e.g. a theater or a meeting), 4) as a passenger in a car for an hour without a break, 5) lying down to rest in the afternoon when circumstances permit, 6) sitting and talking to someone, 7) sitting quietly after a lunch without alcohol and 8) in a car, while stopped for a few minutes in traffic. The range of an item-score is 0 3 on the Likert scale: never doze, slight chance of dozing, moderate chance of dozing and high chance of dozing. The ESS score is the sum of the eight item scores (range, 0 24); higher scores indicate being more sleepy. The Cronbach s alpha value for the ESS in the current study was The ESS has a high sensitivity and high specificity with a cut-off score >10 for daytime sleepiness 11). Those who had an ESS global score of 11 or more are assessed to be impaired in their daily activities by their extraordinary sleepiness, and the remainder are not. To assess not transient but chronic daytime excessive sleepiness, the length of time applied was one month prior to the present study. The other characteristics of sleep habits and sleep problems were also asked of the subjects. Concerning sleep habits, they were asked to respond to the following questions; 1) with respect to your sleep-wake schedules (your bed-time and wake-up time), would you rate it as similar to almost every day (a regular sleep-wake schedule), changes during weekdays compared to weekends (a weekly irregular sleep-wake schedule), changes day by day (a daily irregular sleep-wake schedule), or changes due to shift-work? and 2) would you rate your total sleep time as sufficient (sufficient and adequate) or insufficient (insufficient and very insufficient)? Five items from the PSQI 12, 13) were used for evaluating sleep problems: 1) how many hours of actual sleep did you get per night?; 2) how often have you had trouble sleeping because you cannot get to sleep within 30 min? (DIS); 3) how often have you had trouble sleeping Table 1. Sociodemographic characteristics of 3,909 male white-collar workers Characteristics N (%) Age (years old) (17.0) (22.9) ,215 (30.9) ,178 (29.2) Educational status high school 2,167 (55.6) high school < 1,732 (44.4) Type of work clerk 1,039 (27.0) sales person 1,185 (30.8) researcher 645 (16.8) technical engineer 606 (15.8) manager 193 (5.0) service person 179 (4.7) Marital status married 2,807 (71.9) unmarried 1,099 (28.1) because you wake up in the middle of the night or early morning? (DMS); 4) how often have you had trouble sleeping because you cough or snore loudly?; 5) how often have you taken medicine to help you sleep? (HMU). The subjects were asked to choose one response from the following four answers: not at all, less than once a week, once or twice a week, and three or more times a week, to each question. The presence of symptoms for DIS, DMS, snore and HMU were identified as cases when the subjects responded once or more a week in the previous month to the respective items. The combination of DIS and/or DMS was categorized into difficulty in initiating and/or maintaining sleep (DIMS). Sociodemographic information, physical illnesses, depression, psychoactive substance use affecting sleep such as smoking, drinking alcohol and caffeine-contained beverages were also the items used as putative associated factors in EDS in this study. Depression was measured by means of the CES-D 14, 15). The CES-D yields an item score (range, 0 3) and the sum of the 20-item scores (range, 0 60); higher scores indicate increasing severity of depressive symptoms. It had a Cronbach s alpha value of 0.84 in our present study. This measure differentiates the depressed from the nondepressed well, using a cut-off score of 16 with a high sensitivity and high specificity 16). Smoking status was categorized into non-smoking, less than 1 pack, less than 2 packs and 2 packs or more smoked a day. Alcohol and caffeine consumption were grouped into the quartiles: less than one, less than 4, less than 12

3 Yuriko DOI, et al.: Excessive Daytime Sleepiness among Male White-collar Workers 147 and 12 or more glasses of alcohol consumed a week, and less than 1, less than 3, less than 5 and 5 or more cups of caffeine consumed a day, respectively. Six types of work were grouped into two: clerks and the others 17). Information on detailed occupational class was not obtained in this survey. 3. Statistical analysis Logistic regression analysis was used to examine the association of EDS with its putative factors. A univariate analysis was performed to detect the correlation of EDS with each of the variables, and then the variables were entered to analyze in a multivariate model by the forward selection stepwise procedure (F 0.05 as inclusion and F 0.10 as exclusion). Strong information redundancies were checked and excluded from multivariate logistic regression analysis (e.g., sleep duration and subjective sleep). The variables entered in the models (the reference groups) were as follows: age group (50 59 yr old), educational status (high school or junior high school graduates), marital status (being married), type of work (clerks), sleep duration (7 h of sleep or more per night), sleep-wake schedules (regular), DIMS (less than once a week), snoring (less than once a week), HMU (less than once a week), depression (CESD<16), asthma (absent), peptic ulcer (absent) and muscle-joint pain (absent), smoking (a non-smoker), alcohol drinking (less than one glass a week) and caffeine drinking (less than one cup a day). Subjects with missing values were eliminated from the analyses (listwise exclusion). P values less than 0.05 were considered as significant (two-tailed). The data were analyzed with SPSS version 10.0J 18). Results The prevalence of EDS was 7.2% among male nonshift white-collar workers in the present study. The significant factors associated with EDS in univariate analyses are shown on Table 2. They were as follows: being younger, being unmarried, short sleep duration, insufficient sleep, irregular sleep-wake schedules, having DIMS, snoring, being depressed, having asthma, having peptic ulcer, having muscle-joint pain, and smoking heavily. Subjective sleep was excluded from a multivariate logistic regression analysis because it was closely correlated to sleep duration. Five factors associated with sleep duration, sleep-wake schedules, depression, marital status and smoking still remained significant as P values less than 0.05 after performing a multivariate logistic regression analysis. The odds ratios (95% confidence intervals) were 3.13 ( ) for less than 6 hours of sleep per night, 1.95 ( ) for a daily irregular sleep-wake schedule, 2.96 ( ) for being depressed and 1.48 ( ) for being unmarried. Regarding smoking status, they were statistically marginal when non-smoking was set as a reference: 0.65 ( ), 0.70 ( ) and 2.15 ( ) for less than 1 pack, less than 2 packs and 2 packs or more smoked a day, respectively. Discussion In the present study, we estimated the prevalence rate of EDS as 7.2% among male non-shift white-collar employees working at a telecommunications company in the Tokyo metropolitan area. It was slightly lower than 10.9% for Austrian male workers 19) and higher than 5.0% for Israeli industrial male workers 20). Although variations in prevalence depend on the subjects studied and the questions asked, our finding suggests that EDS should receive attention from the viewpoint of health and occupational safety in the workplace. The investigation of the relationships between EDS and its associated factors is important in order that we develop and propose preventive countermeasures for EDS. We found the factors associated with EDS in a multivariate logistic regression model of our study: short sleep duration, depression, a daily irregular sleep-wake schedule, being unmarried and smoking. The first three factors are consistent with those reported in previous studies on EDS for the general or community population 4, 20 24). It is interesting to note that irregular sleep-wake schedules were significantly correlated with EDS, and a daily irregular sleep-wake schedule affected EDS more strongly than a weekly irregular sleep-wake schedule. This implies that the former type of irregular sleep-wake schedule is much more compromising to the existing circadian sleep-wake rhythms than the latter. In addition, it supports the theory that the transition from internally synchronized to desynchronized rhythms can occur in the multi-oscillator system of the human biological clock, which is remarkable in men, because the period of freerunning circadian rhythms is significantly longer in men than in women 25 27). Since a socially determined sleep-wake schedule is an important factor in the internal synchronization of the human circadian rhythm 27), it is possible that a busy lifestyle (e.g. overtime work or workrelated activities at night) may injuriously affect the sleepwake rhythms of workers. A noteworthy finding in this study was that smoking was identified as a factor associated with EDS in both uni- and multivariate regression analyses (p<0.05). A J- shaped correlation between EDS and smoking status was found: heavy smokers consuming 2 packs or more a day were more likely to feel EDS than non-smokers, whereas smokers consuming less than 2 packs a day were less likely to feel EDS than non-smokers. Considering circadian blood nicotine concentrations during cigarette smoking 28), the combined effect of nicotine stimulant and nicotine withdrawal might produce sleep deprivation and

4 148 J Occup Health, Vol. 44, 2002 Table 2. Factors associated with excessive daytime sleepiness among 3,909 male white-collar workers Univariate Multivariate Factor n OR (95% C.I.) P value OR (95% C.I.) P value Age (years old) ** ( ) ( ) , ( ) , Marital status *** * married 2, unmarried 1, ( ) 1.48 ( ) Sleep duration (hours) *** *** 7 1, , ( ) 1.47 ( ) 6> ( ) 3.13 ( ) Subjective sleep *** Not included sufficient 2, insufficient 1, ( ) Sleep-wake schedule *** * regular 1, weekly irregular 1, ( ) 1.43 ( ) daily irregular ( ) 1.95 ( ) DIMS ** once a week > 2, once a week ( ) Snore * once a week > 3, once a week ( ) Depression *** *** CES-D 16 > 3, CES-D ( ) 2.96 ( ) Asthma * absent 3, present ( ) Peptic ulcer ** absent 3, present ( ) Muscle-joint pain * absent 2, present ( ) Smoking (packs/day) * * 0 2, > ( ) 0.65 ( ) 2> ( ) 0.70 ( ) ( ) 2.15 ( ) DIMS and CES-D mean difficulty in initiating or maintaining sleep and Center for Epidemiologic Studies Depression Scale, respectively. All the variables were entered and analyzed by the forward selection stepwise procedure (F 0.05 as inclusion and F 0.10 as exclusion) in a multivariate model: age group, marital status, educational status, type of work, sleep duration, sleep-wake schedules, DIMS, snore, hypnotic medication use, alcohol, caffeine, smoking, depression, asthma, peptic ulcer and muscle-joint pain. Only the variables with statistical significance on univariate analyses are shown on this table. OR (95% C.I.) means odds ratio (95% confidence interval). *P<0.05, **P<0.01, ***P<0.001.

5 Yuriko DOI, et al.: Excessive Daytime Sleepiness among Male White-collar Workers 149 sleep fragmentation, and consequently daytime somnolence among heavy smokers, whereas the effect on the cholinergic mechanism might produce only daytime arousal among relatively light smokers 29, 30). The prevalence rate of cigarette smoking among men in our study was 46.5%, slightly lower than the Japanese male average (an overall prevalence of 53.5%: 60.9%, 63.4%, 60.0% and 54.1% for those in the 20s, 30s, 40s and 50s, respectively) 31). This might be due to the higher proportion of middle-aged than younger subjects in our study. Strenuous efforts to reduce smoking have been made in workplaces in our country, but they have not succeeded yet. When promoting a smoking cessation program at worksites, our findings suggest that we should consider that smoking impairs not only the physical health of workers but also their daytime alertness at work especially among heavy smokers. Mental heath is one of the critical issues on promoting health at workplaces in our country 32). Sugisawa et al. reported that the prevalence of self-reported mental disorders was 17.5 per 1,000 among 15,639 Japanese male workers aged engaged in various occupations 33). They also reported the standardized prevalence ratios such as 2.94 for computer workers, 1.16 for engineers, 1.06 for clerks, for salesmen and 0.98 for high school teachers. Kawakami et al. focused on examining the prevalence rate of mood disorder in 140 Japanese workers by using the DSM-III-R and estimated 6-month and lifetime prevalence rates of major depressive episodes as 4% and 14%, respectively 34). They suggested no significant difference between white-collar employees and self-employed workers in the prevalence of major depression. We used the CES-D with a cut-off score of 16 and showed the prevalence of depression to be 12.4%. Whether the prevalence in our study subjects is higher or lower than in other studies or different occupational settings is uncertain since the subjects and methods used in the various studies are different. To answer the above question, it is quite sure that a large-scale epidemiological study on mental health including depression should be conducted on workers in the future. It is also necessary in order to prevent or reduce EDS among workers. The limitations of the present study should be addressed. One is the ESS used in our study, a self-report measurement subjectively assessing daytime sleepiness in daily life 10). The subjective measure of the ESS and the objective tool, the Multiple Sleep Latency Test (MSLT), may evaluate different but complimentary aspects of sleepiness 35 38). The findings of our study should therefore be interpreted based solely on subjective aspects of daytime sleepiness measured by the ESS. Such objective assessments as the MSLT and nocturnal polysomnographic recordings could not be incorporated into our occupational based epidemiological study. Further investigation on this issue should be carried out, particularly for the ESS Japanese version. The other limitation was that the subjects studied in our study were restricted to the employees at a certain information and communication technology company in a geographically limited area, so that they were not representative of the Japanese male non-shift white-collar working population. In conclusion, the present study demonstrated the prevalence of and factors associated with EDS among male non-shift white-collar workers in a telecommunications company in the Tokyo metropolitan area. The main findings of this study were that the prevalence rate of EDS was 7.2% and five factors associated with EDS were identified: less than 6 hours of sleep per night, a daily irregular sleep-wake schedule, depression, single status and smoking. These findings imply that the individual behavioral modification of sleep habit and lifestyle is one of the promising strategies for reducing EDS. More importantly, it is possible that occupation policies and health promotions such as a ban on working excessively long hours and a provision of mental health hygiene and social support (e.g. offering psychological counseling and enhancing the bidirectional communications between supervisors and subordinates and among coworkers) at worksites are effective in preventing EDS in the workplace. Acknowledgments: This study was supported by Special Coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science and Technology of the Japanese Government. The authors thank all participants, and the medical and administrative staff of the telecommunications company for their cooperation. References 1) Bonnet M, Arand DL. We are chronically sleep deprived. Sleep 1995; 18: ) Mitler M, Dinges D, Dement W. Sleep medicine, public policy, and public health. In: Kryger M, Roth T, Dement W, eds. Principles and practice of sleep medicine. London: W.B. Saunders, 1994: ) Stoller M. Economic effects of insomnia. Clin Ther 1994; 16: ) Liu X, Uchiyama M, Kim K, et al. Sleep loss and daytime sleepiness in the general adult population of Japan. Psychiatry Res 2000; 93: ) Roth T, Roehrs TA, Carskadon MA, Dement WC. Daytime sleepiness and alertness. In: Kryger M, Roth T, Dement W, eds. Principles and practice of sleep medicine. London: W.B. Saunders, 1994: ) Miteler MM, Miller JC, Lipsitz JJ, Walsh JK, Wylie CD. The sleep of long-haul truck drivers. N Engl J Med 1997; 337: ) Ministry of Health and Welfare, Japan. Society and the low fertility rate. In: Annual report on health and welfare in Tokyo: Ministry of Health and Welfare, Japan, 1998: (in Japanese).

6 150 J Occup Health, Vol. 44, ) Statistics Bureau, Management and Coordination Agency, Japan population census of Japan analytical series No.7-commuting population. Tokyo: Management and Coordination Agency, Japan, 1998 (in Japanese). 9) Obermyer WH, Benca RM. Effects of drugs on sleep. Neurol Clin 1996; 14: ) Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991; 14: ) Johns MW. Sensitivity and specificity of the multiple sleep latency test (MSLT), the maintenance of wakefulness test and the Epworth Sleepiness Scale: failure of the MSLT as a gold standard. J. Sleep Res 2000; 9: ) Buysse DJ, Reynolds CFIII, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res 1989; 28: ) Doi Y, Minowa M, Okawa M, Uchiyama M. Development of the Japanese version of the Pittsburgh Sleep Quality Index. Japanese Journal of Psychiatry Treatment 1998; 13: (in Japanese). 14) Radolff LS. The CESD scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977; 1: ) Shima S, Shikano T, Kitamura T, Asai M. A new selfrating scale for depression. Clin Psychiatry 1985; 27: (in Japanese). 16) Weissman MM, Sholomskas D, Pottenger M, Prusoff BA, Locke BZ. Assessing depressive symptoms in five psychiatric populations: a population study. Am J Epidemiol 1977; 106: ) Uehata T. An epidemiological survey on work, stress and health in various kinds of occupations in Japan, third report. Tokyo, 1996 (in Japanese). 18) SPSS Inc. SPSS 10.0J for windows. SPSS Inc., Tokyo: ) Johns M, Hocking B. Excessive daytime sleepinessdaytime sleepiness and sleep habits of Australian workers. Sleep 1997; 20: ) Lavie P. Sleep habits and sleep disturbances in industrial workers in Israel: main findings and some characteristics of workers complaining of excessive daytime sleepiness. Sleep 1981; 4: ) Breslau N, Roth T, Rosenthal L, Andreski P. Daytime sleepiness: an epidemiological study of young adults. Am J Public Health 1997; 87: ) Hublin C, Kaprio J, Partinen M, Heikkila K, Koskenvou M. Daytime sleepiness in an adult, Finish population. J Intern Med 1996; 239, ) Martikainen K, Urponen H, Partinen M, Hasan J, Vuori I. Daytime sleepiness: a risk factor in community life. Acata Neurol Scand 1992; 86: ) Ohayon MM, Caulet M, Phillip P, Guilleminault C, Priest RG. How sleep and mental disorders are related to complaints of daytime sleepiness. Arch Intern Med 1997; 157: ) Wever RA. Sex differences in human circadian rhythms: intrinsic periods and fractions. Exprimentia 1984; 15: ) Honma K, Honma S, Hiroshige T. Research in human circadian rhythms. Sapporo: Hokkaido University Press, 1989: (in Japanese). 27) Honma K. Human biological clocks and sleep-wake rhythms. Japanese Journal of Medicine 2001; 126: (in Japanese). 28) Benowitz NL, Kuty F, Jacob P 3rd. Circadian blood nicotine concentrations during cigarette smoking. Clin Pharmacol Ther 1982; 32: ) Hughes JR, Gust SW, Skoog K, Keenan R, Fenwick JW. Symptoms of tobacco withdrawal. Arch Gen Psychiatry 1991; 48: ) Prosis GL, Bonnet MH, Berry RB, Dickel MJ. Effects of abstinence from smoking on sleep and daytime sleepiness. Chest 1994; 105: ) 32) Kawakami N, Haratani T. Epidemiology of job stress and health in Japan: review of current evidence and future direction. Ind Health 1999; 37: ) Sugisawa A, Uehata T, He Pin et al.. Mental health, work environment, and health practices among middleaged male workers. Jpn J Ind Health 1993; 35: ) Kawakaimi N, Iwata N, Tanigawa T, et al.. Prevalence of mood and anxiety disorders in a working population in Japan. J Occup Environ Med 1996; 38: ) Johns MW. Daytime sleepiness, snoring, and obstructive sleep apnea-the Epworth Sleepiness Scale. Chest 1993; 103: ) Chervin RD, Guileminault C. Assessment of sleepiness in clinical practice. Nat Med 1995; 1: ) Oslon LG, Cole MF, Ambrogetti A. Correlations among Epworth Sleepiness Scale scores, multiple sleep latency tests and psychological symptoms. J Sleep Res 1998; 7: ) Benbadis SR, Mascha E, Perry MC, Walgamuth BR, Smolley LA, Dinner DS. Association between the Epworth Sleepiness Scale and the Multiple Latency Test in a clinical population. Ann Intern Med 1999; 130: