Sleep disturbance in patients taking opioid medication for chronic back pain

Abstract

Poor sleep is an increasingly recognised problem with chronic pain and further increases the effect on daily function. To identify the relationship between chronic pain, opioid analgesia and sleep quality, this study investigated activity and sleep patterns in patients taking opioid and non-opioid analgesia for chronic back pain. Thirty-one participants (10 healthy controls, 21 patients with chronic pain: 6 on non-opioid medication; 15 on opioid medication) were assessed using actigraphy, polysomnography and questionnaires. Patients with chronic pain subjectively reported significant sleep and wake disturbances as shown by decreased overall sleep quality (Pittsburgh Sleep Quality Index, p < 0.001), increased symptoms of insomnia (Insomnia Severity Index, p < 0.001) and increased fatigue (Fatigue Severity Scale, p = 0.002). They also spent increased time in bed (p = 0.016), took longer to get to sleep (p = 0.005) and had high interindividual variability in other measures of activity but no overall irregular rest-activity pattern. Patients on high doses of opioids (> 100 mg morphine-equivalent/day) demonstrated distinctly abnormal brain activity during sleep suggesting that polysomnography is necessary to detect sleep disturbance in this population in the absence of irregular rest-activity behaviour. Night-time sleep disturbance is common in individuals suffering from chronic pain and may be further exacerbated by opioid treatment. Considerations must be made regarding the appropriate use of combined actigraphy and miniaturised polysomnography for future population-based studies.

Keywords: chronic opioids: side effects; chronic pain; sleep disturbance.

Figure 1

Example actigraphy traces of 12 consecutive days from two controls, two patients with chronic pain who take opioid medication and two patients with chronic pain but on treatment other than opioid medications. Black bars indicate activity, with the same scale used for each actogram. Each actogram is 48 h double‐plotted with successive days on the vertical axis. The midline of each actogram represents midnight between days 1 and 2. In this sample, actigraphy shows continuous activity during the night in patients not on opioid medication (blue bar/arrows) while patients on opioid medication are least active during the night (red bar/arrows). Top left, 62‐year‐old female, top right, 44‐year‐old male. Middle left, 52‐year‐old female morphine equivalent dose = 210 mg, middle right 41‐year‐old male morphine equivalent dose = 71 mg, bottom left, 51‐year‐old female, bottom right 51‐year‐old male. (Opioid dose expressed as morphine equivalents per day).

Figure 2

Top, hypnograms from two healthy controls presenting characteristic electroencephalographic correlates of sleep and its architecture with approximate 90‐min sleep cycles transitioning from wake to NREM light sleep, NREM deep sleep and finally REM sleep. Bottom, hypnograms from two patients taking high doses of opioids (expressed as morphine equivalents per day, MEQ.d⁻¹). Here, abnormal sleep architecture is clearly demonstrated with a large proportion of the night awake or in stage 2 sleep, with striking deficits in REM and deep sleep. The horizontal axis begins at lights off and ends at lights on, with the length dependent upon how long the individual tries to sleep for. The vertical axis corresponds to vigilance state (Wake, REM: Rapid Eye Movement sleep; Stage 1: light sleep; Stage 2: durable sleep; Stage 3: transitional sleep toward deep sleep; Stage 4: deep sleep).

Figure 3

Overnight spectral fast Fourier transform (FFT) of the central C4 electroencephalography channel. (a) healthy control. Patterns are well matched to models of human sleep architecture. Periodic delta frequency bands (0.5–2 Hz) representing deep sleep (stage 3 and 4) are clearly identified in pink, presenting greatest intensity from the beginning of the night and gradually decreasing towards the end. Sigma frequency bands can also be identified (in the range 11–14 Hz) representing sleep spindles of stage 2 sleep. Four bouts of REM sleep can be identified with increasing duration, represented by periods of low intensity and a lack of delta frequencies. (b) Participant on 95 mg of opioids for chronic back pain. In contrast to the healthy control, no discernible pattern of sleep architecture is present. Instead of delta waves, this individual shows theta waves characteristic for stage 2 sleep. Both sigma and delta frequencies are nearly absent, suggesting this individual has abnormal sleep spindles and does not enter deep sleep. Furthermore, there is an absence of REM sleep, with this individual likely alternating between wake, stage 1 and stage 2.

Figure A1

Effect of daily dosage of opioid medication (morphine equivalent, MEQ) on average actigraphic sleep efficiency (left) and sleep fragmentation (right), including and excluding outlier data (n = 14 and n = 13, respectively).