SEARCH

SEARCH BY CITATION

Keywords:

  • chronic sleep loss;
  • introspection;
  • recovery sleep

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

This study addressed a rarely studied question of self-perceptions of performance and overall functional state during cumulative sleep restriction and the ensuing recovery period. Twenty healthy male volunteers, aged 19–29 years, were divided into a sleep restriction group (n = 13) and a control group (n = 7). On the first 2 nights, the sleep restriction group had an 8-h sleep opportunity that was restricted to 4 h for the next 5 nights, and then restored to 8 h for the last 2 nights. The control group had an 8-h sleep opportunity each night. Each day participants accomplished 50-min multitask sessions and gave self-ratings in their connection. Similar to our previous findings on multitasking performance, self-perceived task performance, sleepiness and mental fatigue impaired during the sleep restriction and returned to baseline during the recovery phase. Self-perceived mental effort, tension, task difficulty and task pace showed no sensitivity to the sleep restriction. We concluded that sleep-restricted individuals can probably make use of some self-perceptions when assessing their ‘fitness for duty’. However, at the individual level these measures seem to be inaccurate in revealing actual performance impairments.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Cumulative sleep restriction leads to impairments in objective measures of alertness and cognitive performance (Belenky et al., 2003; Haavisto et al., 2010; Van Dongen et al., 2003). These objective measures are, however, not usually available in the real world, which leaves sleep-restricted individuals with only the method of introspection to determine their ‘fitness for duty’. In this study, the focus will be the question of how sleep-restricted individuals perceive their own performance and functional state.

Most previous studies report that performance self-estimates are affected by sleep loss similarly to actual performance (Baranski and Pigeau, 1997; Baranski et al., 1994; Dorrian et al., 2000, 2007; Jones et al., 2006). In these studies, a variety of performance tasks (e.g. vigilance, mental addition, reasoning and simulated driving tasks) have been used, and participants have been subjected to either sustained wakefulness of up to 64 h or to night shifts before a nighttime testing period. However, also findings on a reduced ability to self-perceive performance impairments during extended wakefulness have been reported (Biggs et al., 2007; Tsai et al., 2005). In these studies, participants either were subjected to only 1 night of partial sleep loss prior to a daytime testing period (Biggs et al., 2007), or their performance self-monitoring was not assessed by true introspective measures, but by behavioral and electrophysiological indices (Tsai et al., 2005). Importantly, most of the above-mentioned findings are based on acute sleep loss experiments. Only the studies of Dorrian et al. (2003, 2007) used a cumulative sleep restriction regimen, but as participants were studied in connection with night shifts it is likely that the results were affected by the circadian factor.

Excluding subjective sleepiness, studies on sleep-restricted individuals' perceptions of their functional state are scarce. A recent study by Odle-Dusseau et al. (2010) is one of the few that has focused on various self-estimates of sleep-restricted individuals in connection with test sessions. The authors found that self-rated effort increased, motivation decreased and stress did not change in connection with two mental tasks under 28 h of sustained wakefulness.

The current study is a continuation of our previous study that showed that both multitasking performance and physiological sleepiness impair over 5 days of sleep restriction and restore during 2 nights of recovery sleep (Haavisto et al., 2010). Now our focus is on the self-ratings that were collected in the same experiment. On the basis of the above-mentioned studies, we expected that performance and sleepiness self-perceptions are sensitive to cumulative sleep restriction and the ensuing recovery period. In addition to these two measures, we examined other pertinent but less-studied self-perceptions of functional state, such as self-perceived mental fatigue, tension, effort and task difficulty, in the same experiment.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Participants

After signing an informed consent, 20 healthy male volunteers (aged 19–29 years) participated in the study. They had a regular sleep–wake pattern (7–9 h per night). Exclusionary criteria included habitual napping, shift work, current or recent medical illnesses, psychiatric illnesses, a history of alcohol or drug abuse, and extreme circadian types. All participants completed the Basic Nordic Sleep Questionnaire (Partinen and Gislason, 1995) and medical screening questionnaires, and they were examined by a physician. The Ethics Committee of the Hospital District of Helsinki and Uusimaa approved the study.

Procedures prior to the experiment

The volunteers adhered to regular sleep schedules for 2 weeks before the experiment. During this period, mean sleep duration measured by a sleep diary and actigraphy was 6 h 53 min (SD = 35.0 min) and 7 h (SD = 51.4 min) in the sleep restriction and control groups, respectively. A few weeks before the experiment, participants had an adaptation night in the laboratory during which a polysomnography was carried out to rule out those with organic sleep disorders.

Design

Participants were randomly assigned to either the sleep restriction group (SR; n = 13) or the control group (n = 7). During the first two experimental days, of which the second served as baseline (BL), the groups had an 8-h sleep opportunity per night (23:00–07:00 h). For the next 5 days (SR1–SR5), the SR group was allowed to sleep for 4 h per night (03:00–07:00 h), after which they had 2 days with 8-h recovery opportunities (23:00–07:00 h; R1, R2). The controls were permitted 8 h of sleep per night (23:00–07:00 h) each day.

Participants accomplished a 50-min multitask session at 10:00, 11:40 and 14:00 h daily. The SR group completed an additional session at 00:30 h. One of the forenoon sessions contained a 10-min rest pause each day. The order of the two types of forenoon sessions was counterbalanced across the days and participants. Participants accomplished other tests during the experiment, but the data from these will not be reported in the current study. The data from the task sessions with the break and conducted at nighttime will be presented elsewhere as well.

Between the test sessions, participants were allowed to watch TV, read and interact with each other. They were under continuous behavioral and electroencephalogram (EEG) monitoring throughout the experiment. Breakfast was served at 07:30 h, lunch at 12:30 h, snack at 15:30 h, dinner at 18:00 h and snack at 21:30 h.

Polysomnography

The sleep EEG data were recorded from the derivations Fp1–A2, Fp2–A1, C3–A2, C4–A1, O1–A2 and O2–A1 using a bandwidth of 0.5–90 Hz and a sample rate of 200 Hz. The recordings were visually scored in 30-s epochs using standard criteria (Rechtschaffen and Kales, 1968).

Multitask

A multitask entitled Brain@Work including four subtasks was used to measure cognitive performance (Fig. 1). It was developed at the Finnish Institute of Occupational Health, and represents a modified version of a multitask entitled SYNWORK (Elsmore, 1994). Below is a description of the four subtasks that were all running in parallel during the test sessions.

image

Figure 1. Example of the user interface of the Brain@Work multitask software. Top left: arithmetic subtask; top right: short-term memory subtask; bottom left: auditory vigilance subtask; bottom right: visual monitoring task.

Download figure to PowerPoint

Short-term memory subtask

Participants first learnt a string of letters. During the test, probe letters were presented at 7-s intervals. The task was to indicate whether or not the probe letter was in the initial string. A correct response gave 10 points, a false response −10 points and a failure to respond −20 points.

Arithmetic subtask

The task was to calculate the sum of two even numbers. The duration of each trial was 7 s. The rules of gaining points were the same as with the short-term memory subtask.

Visual monitoring subtask

The task was to return a moving dot to a circle before it reached the outermost circle. The nearer the dot was to the edge of the circle upon responding, the more points were gained (2, 4, 6 or 10 points). If the dot reached the outermost circle, 10 points per second were subtracted until the dot was returned to the center of the circle.

Auditory monitoring subtask

The task was to discriminate between an infrequent (20% probability), high-frequency (1200 Hz) target tone, and a frequent (80% probability), low-frequency (1000 Hz) non-target tone. The tones were presented randomly at 1.5-s intervals. Points were given in the same manner as in the short-term memory subtasks.

Participants were not shown their points during the experiment. As more than one subtask could be simultaneously active, the task required participants to dynamically allocate attention among the subtasks. Participants trained for the multitask, on average, 245 min before the BL day to flatten the practice effect. To achieve comparable task difficulty among all participants, it was determined individually. For the details of this procedure, see Haavisto et al. (2010).

Introspective measures

Participants rated their sleepiness using the nine-point Karolinska Sleepiness Scale (KSS; Åkerstedt and Gillberg, 1990). Excepting the self-rating scale of performance, all the other self-rating scales were based on nine-point scales, analogous to the KSS. For mental fatigue, the scale varied from ‘feeling very slow, inefficient and lacking initiative’ (1) to ‘feeling very energetic, efficient and full of initiative’ (9); for tension, from ‘feeling very calm and relaxed’ (1) to ‘feeling very tense and strained’ (9); for mental effort, from ‘have not put in any effort at all’ (1) to ‘have put in extremely high effort’ (9); for task pace, from ‘very slow’ (1) to ‘very fast’ (9); for task difficulty, from ‘very easy’ (1) to ‘very difficult’ (9) (Ingre et al., 2002). Only the odd numbers of the nine-point scales were verbally anchored. The self-ratings of sleepiness, mental fatigue and tension were given before and after the task sessions, whereas those of mental effort, task pace and task difficulty only after them. The self-estimates of performance were given before and after the multitask sessions with a visual analog scale that ranged from 100 (100% of the maximum score) to 0 (0% of the maximum score).

Statistical analyses

We used a mixed-model analysis of variance (anova). Our multilevel model included repeated measurements for the same individual. The model also took into account the correlation of the measurements for the same person. The covariance structures in the data were modeled by the compound symmetry method. All explanatory variables were fixed class-variables (factors): group (SR and control); day (BL, SR1, SR2, SR3, SR4, SR5, R1, R2)' and pre–post (pre-task, post-task). The time of the day (forenoon, afternoon) was not included in the final models, as preliminary analyses showed that it did not play a significant role in the ratings. The repeated measurements for each individual were measurements on the day and pre–post factors. The group-variable classes contained different individuals (between-subjects factor). The main focus was on the Group × Day and Group × Day × Pre–Post interaction effects. A separate linear model was conducted to study the accumulation of sleep restriction effects. In this model, the explanatory factors were the same as in the mixed model anova, but now the Day factor was considered as a continuous variable. The estimated slope coefficient derived from the analysis denoted the average daily change. The relationship between self-perceptions and actual performance was studied using Spearman's rank correlation coefficients. The data were analysed using pro mixed in sas 9.2 (SAS Institute Inc., Cary, NC, USA, 2009).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Sleep

The SR group slept, on average, 7 h 19 min (SD = 17.4 min) during the baseline night, 3 h 52 min (SD = 2.4 min) a night during the SR phase, and 7 h 40 min (SD = 8.3 min) a night during the recovery period. The control group averaged 7 h 20 min (SD = 13.4 min) of sleep per night.

Self-perceived performance

Of the interaction effects, only the Group × Day was significant (F7,126 = 6.61, < 0.0001). Planned comparisons showed that the groups differed from each other on SR4 (F1,126 = 3.98, P = 0.0482) and SR5 (F1,126 = 7.88, P = 0.0058; Fig. 2; Table 1). In the SR group, the pre-task and post-task ratings lowered 8.7%-units and 17.2%-units across BL–SR5, respectively, whereas the corresponding drop was 23.0%-units in the actual performance level (1%-unit equaled 128 points). The estimated slope coefficient derived from the linear model analysis was −1.6423 (t = −3.58, P = 0.0004) for the pre-task ratings, and −2.7544 (t = −6.00, < 0.0001) for the post-task ratings across BL–SR5 in the SR group. In the control group, the corresponding coefficients (pre-task: 0.8521; post-task: 0.7888) remained non-significant. The relationship between self-perceived and actual performance across BL–R2 varied greatly within the SR group (Fig. 3; Table 2). The Spearman's correlation coefficients calculated separately for each individual varied between −0.40 and 0.98 (mean 0.60) for the pre-task ratings, and between 0.37 and 1.00 (mean 0.68) for the post-task ratings.

Table 1. Means (and SEM) of the pre- and post-task self-estimates during the experimental days in the SR and control groups
Self-estimateBaselineSR1SR2SR3SR4SR5R1R2
Pre-Post-Pre-Post-Pre-Post-Pre-Post-Pre-Post-Pre-Post-Pre-Post-Pre-Post-
  1. Baseline, baseline day; SR1–SR5, 1st–5th sleep restriction days; R1–R2, 1st–2nd recovery days; Pre-, pre-task estimate; Post-, post-task estimate.

  2. For all estimates, the scale ranges from 1 (low) to 9 (high), except for mental fatigue for which the nine-point scale is reverse (i.e. 1 denotes ‘high’ and 9 ‘low’).

Sleepiness
SR group3.6(0.3)5.1 (0.4)4.0 (0.2)5.8 (0.4)4.3 (0.3)6.7 (0.4)4.6 (0.3)7.0 (0.4)5.0 (0.3)6.7 (0.4)5.4 (0.4)6.9 (0.5)3.9 (0.3)5.0 (0.6)3.3 (0.4)4.3 (0.5)
Control3.6 (0.2)5.3 (0.3)3.8 (0.2)5.4 (0.4)3.8 (0.4)5.6 (0.5)4.1 (0.4)5.0 (0.4)3.9 (0.3)5.1 (0.3)4.1 (0.4)5.3 (0.4)4.0 (0.4)5.0 (0.5)3.9 (0.3)4.5 (0.3)
Mental fatigue
SR group6.0 (0.3)5.0 (0.3)5.6 (0.2)4.2 (0.3)5.2 (0.3)3.9 (0.3)4.8 (0.2)3.8 (0.3)4.7 (0.3)4.0 (0.3)4.7 (0.3)4.0 (0.4)5.8 (0.3)5.3 (0.4)5.9 (0.3)6.1 (0.4)
Control6.0 (0.2)5.0 (0.4)6.1 (0.3)5.0 (0.4)6.0 (0.4)4.5 (0.4)5.8 (0.3)5.4 (0.3)5.9 (0.3)5.2 (0.4)6.1 (0.3)5.1 (0.5)5.9 (0.4)5.4 (0.4)6.0 (0.4)5.7 (0.3)
Tension
SR group3.0 (0.2)4.1 (0.3)3.6 (0.3)4.2 (0.3)3.7 (0.3)4.3 (0.3)3.8 (0.2)4.3 (0.3)3.7 (0.2)3.7 (0.4)3.6 (0.3)4.3 (0.3)3.5 (0.2)3.7 (0.4)3.7 (0.3)3.8 (0.3)
Control3.5 (0.3)4.8 (0.5)3.9 (0.4)4.8 (0.4)3.9 (0.4)4.7 (0.6)3.8 (0.4)4.6 (0.6)3.9 (0.4)4.4 (0.6)3.8 (0.3)4.5 (0.5)3.9 (0.5)4.8 (0.6)3.8 (0.4)4.4 (0.6)
Effort
SR group 6.3 (0.3) 6.0 (0.3) 5.8 (0.3) 5.7 (0.3) 6.3 (0.1) 5.7 (0.3) 6.0 (0.3) 6.5 (0.3)
Control 6.6 (0.3) 6.5 (0.5) 5.9 (0.6) 6.3 (0.4) 6.4 (0.5) 5.8 (0.4) 5.6 (0.6) 6.2 (0.4)
Task pace
SR group 5.3 (0.3) 5.2 (0.3) 5.5 (0.3) 5.6 (0.2) 5.8 (0.3) 5.6 (0.3) 5.0 (0.2) 4.7 (0.3)
Control 6.0 (0.4) 5.8 (0.3) 6.0 (0.5) 5.9 (0.5) 5.8 (0.5) 5.6 (0.4) 5.6 (0.4) 5.6 (0.3)
Task difficulty
SR group 5.0 (0.3) 5.2 (0.3) 5.3 (0.3) 5.3 (0.3) 5.4 (0.3) 5.2 (0.3) 4.9 (0.3) 4.2 (0.3)
Control 5.6 (0.2) 5.9 (0.4) 6.0 (0.5) 5.6 (0.5) 5.6 (0.5) 5.6 (0.5) 5.9 (0.6) 5.4 (04)
Table 2. Individual-level correlations between actual multitasking performance and the self-perceptions of performance and sleepiness across the experimental days (BL–R2) in the SR group
ParticipantMultitasking performance and pre-task performance self-ratingsMultitasking performance and post-task performance self-ratingsMultitasking performance and pre-task sleepiness self-ratingsMultitasking performance and post-task sleepiness self-ratings
  1. Participants have been arranged by the degree of actual performance impairment across the SR days (individual with the largest impairment first).

  2. *< 0.05, **P < 0.01, ***< 0.001.

10.710.74*−0.86**−0.84**
20.81*1.00***−0.58−0.69
30.81*0.78*−0.66−0.72*
40.600.88**−0.60−0.85**
50.88**0.80*−0.98***−0.93***
6−0.400.76*−0.66−0.76*
70.79*0.43−0.67−0.59
80.480.37−0.28−0.63
90.71*0.62−0.48−0.64
100.82*0.90**−0.76*−0.59
110.160.44−0.56−0.60
120.490.53−0.45−0.21
130.98***0.57−0.83*−0.88**
image

Figure 2. The daily means (and SEM) of self-perceived performance (% of the maximum) rated before (solid black line with squares) and after (solid gray line with circles) each 50-min multitask session. The daily means (and SEM) of actual multitask performance (total score obtained) are also shown (dotted black line with triangles). BL, baseline; SR1SR5, 1st5th sleep restriction day; R1–R2, 1st2nd recovery day.

Download figure to PowerPoint

image

Figure 3. Self-perceived and actual performance presented as standardized z-scores for each sleep-restricted individual across BL–R2. Participants have been arranged by the degree of actual performance impairment following the sleep restriction regimen (individuals with the largest impairments on the top). Pre-task ratings: solid black line with squares; post-task ratings: solid gray line with circles; actual performance: dotted black line with triangles. The y-axis scale has been individually adjusted for each participant.

Download figure to PowerPoint

Self-perceived sleepiness

Of the interaction effects, all two-way interactions became significant (Group × Day: F7,126 = 7.21, < 0.0001; Group × Pre–Post: F1,18 = 4.57, P = 0.0472; Day × Pre–Post: F7,126 = 2.23, < 0.0359). Planned comparisons showed that the sleep-restricted participants rated themselves sleepier than the controls on SR3 (F1,126 = 5.83, P = 0.0172), SR4 (F1,126 = 6.88, P = 0.0098) and SR5 (F1,126 = 8.52, P = 0.0042; Table 1). The slope coefficients derived from the linear model analysis revealed that the average daily change was 0.3477 units (t = 7.09, < 0.0001) for the pre-task ratings, and 0.3534 units (t = 7.21, < 0.0001) for the post-task ratings across BL–S5 in the SR group. In the control group, the corresponding coefficients (pre-task: 0.1033; post-task: −0.05906) were not significant. The relationship of self-perceived sleepiness to actual performance across BL–R2 varied markedly in the SR group (Table 2). The Spearman's correlation coefficients calculated separately for each individual ranged between −0.28 and −0.98 (mean −0.64) for the pre-task ratings, and between −0.21 and −0.93 (mean −0.69) for the post-task ratings.

Self-perceived mental fatigue

The Group × Day and Day × Pre–Post interaction effects reached significance (Group × Day: F7,126 = 6.19, < 0.0001; Day × Pre–Post: F7,126 = 2.89, P = 0.0079). Planned comparisons showed that the sleep-restricted participants rated themselves sleepier than the controls on SR3 (F1,126 = 10.19, P = 0.0018), SR4 (F1,126 = 8.56, P = 0.0041) and SR5 (F1,126 = 9.56, P = 0.0024; Table 1). The linear model analysis showed that the slope coefficient was −0.2597 (t = −5.79, P < 0.0001) for the pre-task estimates, and −0.1721 (t = −3.83, P = 0.0001) for the post-task estimates across BL–SR5 in the SR group. The corresponding coefficients (pre-task: −0.01764; post-task: 0.07815) were not significant for the controls.

Self-perceived tension

Only the effects of the Pre–Post factor (F1,18 = 53.15, < 0.0001) and the Group × Pre–Post interaction reached significance (F1,18 = 4.45, < 0.0492). Participants rated themselves more tensed post-task than pre-task, and this difference was more pronounced for the control than the SR group (Table 1).

Self-perceived effort

The Group × Day interaction effect was not significant. The self-ratings showed that both groups put quite a lot of effort into their task performances each day (Table 1).

Self-perceived task pace and difficulty

Neither the self-ratings of task pace nor those of task difficulty were affected by the Group × Day interaction. Participants experienced the task rather average in terms of these characteristics (Table 1).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

Our results show that the self-ratings of multitasking performance, sleepiness and mental fatigue are affected by cumulative sleep restriction and the ensuing recovery sleep period. No evidence was found to suggest that sleep restriction would affect the self-ratings of tension, effort, task difficulty and pace. In all, these findings suggest that: (i) the effects of cumulative sleep restriction on introspective measures vary greatly from one measure to another; and that (ii) those introspective measures that are affected restore after the first recovery sleep opportunity.

Introspective measures showing evidence of sensitivity to cumulative sleep restriction

Self-perceived performance

Our finding that cumulative sleep restriction similarly affects self-perceived and actual performance is in line with most of the findings from acute sleep deprivation studies (Baranski and Pigeau, 1997; Baranski et al., 1994; Dorrian et al., 2000, 2007; Jones et al., 2006). This suggests that both types of sleep limitations result not only in similar impairments in task performance (Van Dongen et al., 2003), but also in similar changes in performance self-ratings.

Though the effect of cumulative sleep restriction was observed for the objective and introspective measures of performance, the correlation between the measures varied greatly among the sleep-restricted individuals and, on average, it was not strong. This observation accords to the findings of previous works that the association between these two measures is moderate at best under sleep loss (Dorrian et al., 2003, 2007).

In spite of the fact that changes in self-estimated and actual performance induced by sleep loss do not perfectly match, performance self-perceptions can be considered as potential measures to improve safety under sleep restriction conditions (Dorrian et al., 2000). The next steps would be to know to what extent individual differences in performance response to sleep restriction are manifested in self-estimates, and which are the best ways of measuring them. It may be essential that sleep-restricted performers have several moments to estimate their ongoing performance. In addition, performance feedback and concentration on the most critical aspects of performance may improve possibilities to use self-perceived performance as a proxy for actual performance (Dorrian et al., 2003, 2007).

Interestingly, most of the restoration process occurred during the first recovery sleep opportunity. A similar pattern was observed for multitasking performance in our previous paper, albeit the process continued during the second recovery night (Haavisto et al., 2010). This similitude between self-perceived and actual performance supports the idea of using self-perceptions, not only to assess performance impairments but also to assess their recovery.

Self-perceived sleepiness

Our previous study already showed that subjective sleepiness increased in the course of cumulative sleep restriction and restored during the recovery phase (Haavisto et al., 2010). Similar findings have also been found in other studies, albeit the results on recovery pace vary (Banks et al., 2010; Belenky et al., 2003; Dinges et al., 1997).

High mean sleepiness levels (approximately 7 on the KSS) were observed only post-task during the last three SR days. Interestingly, sleepiness was clearly at lower levels (approximately 5 on the KSS) in the beginning of these task sessions. In the real world, mild pre-task sleepiness may make it difficult for sleep-restricted individuals to see that it can be hazardous to start a driving task, for example. This argument is based on the fact that once an individual has started performing it is often difficult for him or her to interrupt it because a pause may have negative consequences, such as increased time pressure or disengagement from the task (Jett and George, 2003).

Similar to self-perceived performance, subjective sleepiness restored after the first recovery night. This pace is similar to the findings of Belenky et al. (2003). However, Banks et al. (2010) reported that subjective sleepiness did not completely restore after a 8–10-h recovery sleep opportunity following 5 days of sleep restricted to 4 h a night. There are at least two differences that could explain this discrepancy. Banks et al. measured sleepiness between 08:00 and 20:00 h, while we measured between 09:00 and 14:00 h. It is possible that a residual effect of sleep restriction is manifested more clearly during the later part of the day than during the earlier one. Also, Banks et al. measured the post-task KSS ratings after 13 min of performance, while we measured the same after 50 min of performance. Having such a long performance period behind when rating sleepiness may mask possible minor differences in sleepiness that are measurable when the preceding task session is markedly shorter.

Interestingly, self-perceived sleepiness and self-perceived performance correlated almost equally strongly with actual performance. This finding suggests that changes in actual performance levels during sleep restriction and the following recovery period can be assessed with similar accuracy by both measures.

Self-perceived mental fatigue

The sensitiveness of mental fatigue to sleep restriction found in the current study is in line with the results of Ingre et al. (2002). Also, Dinges et al. (1997) found that mental exhaustion and vigor impaired in the course of a 7-day sleep restriction regimen. Moreover, a recent population-based study revealed that self-rated presenteeism was increased among those with a sleep disorder (Swanson et al., 2011). These findings support the hypothesis that chronically insufficient sleep leads to decreased mental energy. Similar to perceived sleepiness and performance, the elevated levels of mental fatigue were restored following one 8-h sleep opportunity.

Introspective measures showing no evidence of sensitivity to cumulative sleep restriction

Self-perceived mental effort, task difficulty and task pace

Our finding of the insensitivity of self-rated effort to sleep restriction is in line with studies by Engle-Friedman et al. (2003) and Drummond et al., 2000;. On the other hand, some studies have reported on increased mental effort under sleep loss (Dinges et al., 1992; Odle-Dusseau et al., 2010; Pilcher and Walters, 1997). In addition to mental effort, self-perceived task difficulty and pace showed no sensitivity to sleep restriction. Together these findings propose that sleep-restricted individuals, whose performance is clearly impaired and who perceive this pattern to at least some degree, do not necessarily experience their task more difficult or faster or put more effort into their performance. An interesting question for the future studies is whether this phenomenon also exists in the real world, where a certain performance level is often required to avoid losses and damages.

Self-perceived tension

Our result on the insensitivity of self-perceived tension to sleep restriction is understandable, considering that self-perceived task difficulty, task pace and effort did not change either. In addition, participants' awareness of receiving no consequences for performance errors is probably a main reason for not seeing the task stressful despite degraded performance. One explanation for our finding may also be that it is contradictory to feel oneself hyperaroused (stressed) simultaneously with feeling hypoaroused (drowsy), as these two represent the opposite ends of the arousal continuum.

Study limitations

Our sample consisted of young healthy men, which limits the application of the results to other age groups, women, unwell people, and to those who are experienced in performing while sleep restricted. In addition, the small sample size may explain why some subjective measures showed no effects of sleep restriction. Second, the study was conducted in laboratory conditions, in which performance deteriorations had no real consequences for the performer or others. In the future, a similar study should be conducted in a real world setting. Also, we did not test whether sensitivity to sleep restriction was increased in the SR group after the recovery period. It is possible that if the recovery phase had been followed by a second SR period some residual effect from the first SR period would have come out. Finally, the results may be affected by methodological limitations. The performance self-rating scale was quite general by nature and presented quite infrequently during the task sessions. Secondly, most of the other self-rating scales were not well established, and thus the findings based on them may underestimate the true effect of sleep restriction.

Conclusions

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

The current study proposes that sleep-restricted individuals can probably make use of some self-perceptions, such as self-perceived performance, sleepiness and mental fatigue, when assessing their ‘fitness for duty’. However, it is important to keep in mind that at the individual level these measures seem to be inaccurate in revealing actual performance impairments. Moreover, there seem to be introspective measures, such as self-perceived effort, tension, task pace and task difficulty that are quite resistant to the effects of sleep restriction. The recovery process of those introspective measures that are affected by sleep restriction may be even faster than that of impaired performance.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References

The authors thank the nursing and technical staff of their laboratory for their contribution to the collection and analysis of the data. The study was supported by grants from the Finnish Work Environment Found and the National Technology Agency of Finland.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgements
  9. References