Occupational Medicine & Health Affairs
Open Access

Obstructive sleep apnea hyponea syndrome (OSAHS) is a common disease with reported rates of 3-7% in middle age population [1]. It’s characterized by repetitive partial or complete collapse of the airway during sleep, resulting in apnea or hypopnea. These episodes cause intermittent hypoxia, repetitive arousal from sleep, which leads to increased sympathetic activity, increased respiratory effort, and sleep fragmentation. Untreated OSAHS is incriminated in increased risk of several co-morbidities including cardiovascular diseases, and metabolic syndrome. Disrupted sleep can also result in tiredness, lack of energy, and a feeling of weakness during the day. Severe forms of the disease are often associated with excessive daytime sleepiness and micro-sleep episodes. Furthermore, persons with OSAHS may experience decreased quality of life, significant cognitive limitations and mood disorders such as anxiety, irritability, and depression.

Because of these adverse mood changes and neurocogintive dysfunction, it has been suggested that OSAHS may be detrimental to work performance, and associated with greater risk of workplace accidents. However, this hypothesis has been little explored, and even less so among North African population and no solid conclusion could be drawn due to disputed results.

Considering the above findings and rationale, we designed an observational cross-sectional study in order to gain a greater understanding in how OSAHS and daytime sleepiness affect work productivity and workplace accidents in a sample of subjects referred for suspected sleep-disorders.

Participants

The study was performed during the period of July 2015 to April 2017 on a subset of participants referred to the department of respiratory and sleeps medicine of Hedi Chaker University Hospital in Sfax, Tunisia for suspected sleep-disorders. Only subjects between the age of 18 and 60 years, and actively employed for at least 6 months were included in the study. The exclusion criteria were: another sleep disorder, previous treatment for sleep apnea, medications that could affect sleep and/or vigilance, serious medical conditions, overt psychiatric disorders including alcohol or other drug addiction, and pregnancy. The study was approved by the institutional Human research ethics committee, and consent was obtained from all participants.

Instruments

Self-administrated questionnaires: Consenting participants were given extensive questionnaire regarding their socio-demographics, habits, medical history, and working conditions such as work schedule, shift work, number of worked hours and the kind of work they had over the last 24 months. The following groupings were classified as white-collar workers: clerks, technicians and associate professionals, highly skilled professionals, directors and managers. Otherwise, unskilled workers, production-line/machine workers and drivers, skilled manual workers, farm/horticulture workers, service workers and retailers were classified as blue-collar workers [2]. The survey also included questions aimed at evaluating the number of work-related accidents, the number of sick leave more than 30 days over the last year, as well as somatic symptoms of stress such as epigastralgia, chest tightness, dyspnea and muscle tension.

Elsewhere, participants completed the Work Productivity and Activity Impairment (WPAI) questionnaire, in order measure impairments in work and activities. Daytime sleepiness was assessed by Epworth Sleepiness Scale (ESS), while Patient Health Questionnaire (PHQ9) was used to screen for the presence and severity of depression according to DSM-IV criteria.

Work productivity and activity impairment (WPAI) questionnaire: The WPAI [3] is a validated instrument that consists of six questions: Q1=currently employed; Q2=hours missed due to health problems; Q3=hours missed due to other reasons; Q4=hours actually worked; Q5=degree health affected productivity while working; and Q6=degree health affected regular activities. According to the questionnaire responses, scores are generated for four main outcomes including absenteeism (work time missed), presenteeism (impaired productivity at work), and overall work productivity loss (absenteeism +presenteeism) in the past 7 days. Impairment in performing daily (non-work) activities, such as work around the house, shopping, child care, exercising, etc. is also assessed.

The different outcomes are expressed as percentage score, with higher scores indicating greater impairment. They are generated as the following:

1. Absenteeism = Q2/(Q2+Q4) for those who were currently employed;

2. Presenteeism = Q5/10;

3. Percent overall work loss due to health= Q2/(Q2+Q4)+[(1-Q2/ (Q2+Q4)) × Q5/10] for those who were currently employed;

4. Percent activity impairment due to health= Q6/10 for all respondents. For those who missed work and did not actually work in the past seven days, the percent overall work impairment due to health is equal to the percent work time missed due to health.

Patient health questionnaire (PHQ9): The PHQ-9 is a validated selfadministered tool used in primary care to screen patients for depression and to estimate symptoms severity. By assessing patient’s mood over the last 2 weeks, the questionnaire allows scoring each of the 9 DSM-IV criteria for Major Depressive Disorder as ‘0’ (not at all) to ‘3’ (nearly every day). As a severity measure, the PHQ-9 score can range from 0 to 27. PHQ-9 scores of 5, 10, 15, and 20 represented mild, moderate, moderately severe, and severe depression, respectively [4].

Epworth sleepiness scale (ESS): Daytime sleepiness was estimated using the Arabic version of the ESS (AESS) [5]. It’s an eight selfcompleted questionnaire assessing the probability of falling asleep in various daytime situations, for a total score ranging from 0 to 24. Depending on their ESS score (ESS<11, ESS ≥ 11), subjects were divided into two groups of subjective sleepiness severity. Those with an ESS score ≥ 11 were considered to have significant daytime sleepiness.

The AESS was created according to the recommendations of the ISPOR Task Force for Translation and Cultural Adaptation with bilingual individuals (Table 1). It had shown a high internal consistency (Cronbach’s alpha 0.76, intra-class correlation coefficient of 0.85 (IC95%: 0.76–0.92), thus it’s considered valid instrument for the evaluation of daytime sleepiness in Arabic speaking populations [5].

Table 1: Socio-demographic and clinical characteristics of study participants.

Nocturnal polygraph recording: Nocturnal polygraph recording was performed for all patients using type III polysomnography (PSG) (EmblaTM N7000, ResMed Corp., San Diego, CA, United States) according to standard techniques. Airflow was monitored by an oronasal air pressure transducer (EmblaTM N7000, ResMed Corp., San Diego, CA, United States), thoracic and abdominal respiratory effort and body position by uncalibrated impedance plethysmography belts (EmblaTM N7000, ResMed Corp., San Diego, CA, United States), and arterial oxygen saturation (SpO₂) by digital pulse oximetry. Snoring was recorded by a microphone (EmblaTM N7000, ResMed Corp., San Diego, CA, United States) placed on the anterior neck. Prior to sleep study, patients were recommended to ovoid caffeinated drinks and foods, alcohol, and drugs that could interfere with sleep architecture.

Scoring respiratory events was performed manually by one doctor according to the standard criteria of the American Academy of Sleep Medicine (AASM) published in 2012 [6]. Obstructive apnea was defined as a 90% decrease in airflow relative to the basal amplitude persisting for at least 10 seconds in the presence of thoracic and abdominal movements. Hypopnea was defined as a 30% decrease in the airflow amplitude relative to the baseline persisting for at least 10 seconds, with an associated 3% oxygen desaturation. Patients having an apnea hypopnea index (AHI) <5/hour without any evocative symptoms apart from snoring were classified as controls, while patients with an AHI of 5 or greater and clinical symptoms were considered positive for OSAHS [7].

According to the AHI, subjects were classified into groups. One group included subjects with severe OAHS (AHI ≥ 30), while a second included subjects without OAHS (

Depression rates: Considering the total sample, PHQ9 score was significantly higher in subjects with daytime sleepiness compared with those with normal ESS (p=0.0001). The observed difference between the two groups was significant whatever the AHI was Table 2). In contrast, no significant difference was found in PHQ9 score according to AHI, T90%, and IDO.

Table 2: Assessment of depression severity with patient health questionnaire according to Epworth sleepiness scale.

Somatic symptom of stress, and work related accidents: Somatic symptom of stress, near miss accidents in the workplace, and work related accidents were not associated with any sleep polygraphic parameters including AHI, ODI, and T90% (Figure 1). However, near miss accidents in the workplace and work related accidents were significantly more frequent in sleepy subjects (ESS ≥ 11), as compared with ESS

Figure 1: Study schema and a priori sample quota algorithm. WPAI, work productivity and activity impairment questionnaire. AHI, apnea-hypopnea index. ESS, epworth sleepiness scale.

Work productivity: Among the four outcomes of the WPAI scale, only presenteeism was associated with AHI. Actually, patients with AHI ≥ 30 had a significantly higher score of presenteeism in comparison with those with AHI Figure 2).

Figure 2: Relationship between work productivity and activity impairment questionnaire (WPAI) sores, and apnea-hypopnea index (AHI). The means for each WPAI outcome are calculated for two levels of epworth sleepiness scale (ESS) (

Elsewhere, when most hypoxemic subjects (T90% ≥ 30%) were compared with the least hypoxemic subjects (T90% <30%), significant higher presenteeism was observed (46.67 ± 28.08 vs. 31.57 ± 27.57; p=0.032).

Considering subjective sleepiness, presenteeism, impairment in performing daily activities and overall work productivity loss scores were significantly higher in sleepy subjects when compared with nonsleepy ones. Absenteeism was similar in both groups. These results were observed in the sample as a whole as well as among the two AHI groups (Figure 3).

Figure 3: Relationship between work productivity and activity impairment questionnaire (WPAI) scores, daytime sleepiness, and apnea hyponea index. The means for each WPAI outcome are calculated for two levels of epworth sleepiness scale (ESS) (

A similar pattern was observed when subjects were divided into blue and white-collar workers with significant higher scores in sleepy subjects for presenteeism, impairment in performing daily activities and overall work productivity loss, but not for absenteeism. It is interesting to observe, however, that the impact of daytime sleepiness was more marked among white-collar workers (Figure 4).

Figure 4: Relationship between work productivity and activity impairment questionnaire (WPAI) scores, daytime sleepiness, and type of work. The means for each WPAI outcomes are calculated for two levels of epworth sleepiness scale (ESS) (

In the total sample, a weak correlation was found between presenteeism and both AHI (r=0.212, p=0.012) and T90% (r=0.222, p=0.009) (Table 3). In addition, the four outcomes of the WPAI scale were weakly correlated with PHQ9 score. Weak correlations were also detected between both presenteeism and overall work productivity loss scores and ESS. This correlation was stronger among white-collar workers (r=0.624, p=0.0001; r=0.602, p=0.0001 respectively). In the same group, presenteeism and overall work productivity loss scores were moderately correlated with AHI (r=0.395, p=0.025; r=0,382, p=0.031 respectively).

Table 3: Correlations of work productivity and activity impairment scores with apnea hypopnea index, patient health questionnaire and epworth sleepiness scale.

In this study, a decreased work performance (as assessed by a validated scale) was observed in patients with severe OSAHS, but also and most importantly in those with daytime sleepiness. This impact has been found to be more marked in white-color workers as compared with blue-color workers.

Despite the impact of OSAHS on occupational health has been suspected since a long time, only few studies have been performed to evaluate this relationship, showing divergent results. Jurado-Gimez B et al. [8] studied a sample of 259 Spanish individuals referred to the sleep disorders clinic for suspected OAHS. Participants were classified to OAHS group or a non-OAHS according to polysomnography results. Both groups had similar rates of work-related accidents. However, OAHS was significantly associated with higher rate of absenteeism. Psychological distress and OAHS were associated with significantly poorer work productivity as assessed on the Spanish IMPALA (impact of disease on work productivity) index. A diagnostic of OAHS was the variable with the most importance impact on the working lives of patients, while excessive daytime sleepiness and subjective quality of sleep (as assessed using Pittsburg Sleep Quality index) was not associated with any influence.

For his part, Muglgrew et al. [9] studied a sample of 428 Canadian adults that had undergone overnight polysomnography for sleepdisordered breathing. OAHS severity defined by AHI was not associated with any dimension of work limitation questionnaire (WLQ). In contrast, subjective sleepiness as assessed by ESS was strongly associated with three of the four scales of WLQ.

Similarly in our study, work performance as assessed by WPAI was significantly affected in sleepy patients with ESS ≥ 11, when compared with non-sleepy patients, while the impact of OAHS severity defined by AHI was less clear. The same conclusion was reached by Bean [10] when comparing a group including 1759 subjects with OAHS, depression, narcolepsy, multiple sclerosis, with 1977 subjects without these conditions. In both, group, excessive sleepiness was associated with highly significant impairment in work productivity for all measures, except for absenteeism.

Many other studies had also reached the same conclusion about the negative impact of daytime sleepiness on work performance [11-13] however some of these have limitations. Actually, surveys were almost always focused on subjects referred for suspected-disordered breathing rather a community-based sample as was the case for Muglgrew et al. [9]. No objective evaluation was performed to better assess the impact of daytime sleepiness on work performance. Some authors did not perform an objective sleep study [11,13] while others used only polygraph as their sleep study [14].

Yet, despite the above limitations, we could still deduce that daytime sleepiness as a common feature of OAHS could cause greater impairment than AHI on work performance. These findings would support the current controversy about the value of AHI as an indicator of OAHS severity. Actually, several recent studies have suggested that AHI would not be enough to predict adverse safety and health consequences of OAHS, particularly daytime fatigue and sleepiness [15]. This is consists with the results of our study that showed no association between ESS and AHI. Therefore, patients with mild to moderate disease (5 ≤ AHI <30) may be more tired and sleepy, and consequently have lower work performance than patients with severe disease (AHI ≥ 30). This means that AHI would not be appropriate to objectively assess the impact of OAHS on work performance. Thus, a novel parameter such as the recently suggested AHI adjusted should be used to improve the recognition of OAHS patients, as well as accuracy and reliability of disease severity estimates [16].

In our study, depression score was not associated with any polygraphic parameter, but was significantly higher in patients with ESS ≥ 11. This fact may contribute to impair work performance in sleepy patients. Elsewhere, Mulgrew et al. found that the impact of subjective daytime sleepiness as assessed by ESS on work performance was equal between white-color and blue-color workers [17]. As opposed to these findings, white-color workers appeared to be more marked affected in our study. This difference may be explained by the sedentarily nature of white-color work that makes sleepiness harder to fight off.

Some studies have suggested an increased risk of work-related accidents due to OAHS symptoms. However, only very few studies have assessed this relationship in patients with definitive diagnosis of OAHS. Besides, studies did not consistently find that daytime sleepiness and the severity of OAHS were correlated with crash risk [8,17]. Our results showed that Work-related accidents were associated with ESS, but not with AHI. In line with this study, Lindberg et al. [18] showed a 2-fold increase in the risk of work-related accidents in snorer and sleepy subjects when compared with subjects without those symptoms. Jurado-Gimez B et al. [8] reported that the OAHS group and the control group had similar rates of workplace accidents.

Our study has limitations. First, the sample would not be representative of our target population as it was small in size, and was selected among subjects referred for suspected sleep disordered breathing. This may have introduced a selection bias making our results not representative enough to be extrapolated beyond our referral group. A second limitation may be related to using polygraphy, and not polysomnography, for diagnosing OAHS which may lead to an underestimation of the prevalence and the severity of the disease. Third, work performance was assessed by WPAI scale which has been validated into many languages, but neither in Arabic nor in Tunisian dialect. In addition, despite many confounders were considered when including subjects, it remains still difficult to conclude about the association of work performance with OAHS and daytime sleepiness. The contribution of other unrecognized confounders in this association could not be excluded given the cross-sectional nature of our study. Finally, there was no follow up-analysis of patients posttreatment which would be helpful to have more information about the association in question.

To our knowledge, this is the first comprehensive study of the burden of OSAHS and daytime sleepiness on occupational health in a Nord African population. Our study demonstrated that daytime sleepiness may lead to low work performance in a population referred for suspected sleep-disordered breathing. This impact was more marked in white-color workers as compared with blue-color workers. In contrast, the relationship between OAHS (as defined by AHI) and work performance was less clear. These results suggest that screening for excessive daytime sleepiness should be considered as a part of occupational health assessment. It would be also interesting to define new instruments other than conventional AHI to more accurately evaluate the impact on the working lives.