Author Year

Country

Study Design

Sample Size

Grima et al. (2018)

Australia

RCT Crossover

PEDro=9

N=33

Population: Melatonin-placebo group (N=18): Mean Age=35yr; Gender: Male=61%, Female=39%; Median Time Post Injury=61mo; Median GCS= 5. Placebo-melatonin group (N=15): Mean Age=38yr; Gender: Male=73%, Female=27%; Median Time Post Injury= 25mo; Median GCS=8.

Intervention: Participants with chronic insomnia were randomly allocated to a 4wk melatonin or placebo treatment before crossover. Melatonin formula was a prolonged release formula (2mg). Participants were measured at baseline and at the end of each treatment phase.

Outcomes:  Pittsburgh Sleep Quality Index (PSQI); Sleep onset latency (measured by wrist actigraphy); Epworth Sleepiness Scale (ESS); Hospital Anxiety Depression Scale (HADS); Fatigue Severity Scale (FSS); Short-form health survey (SF-36 v1) subscales: Physical functioning (PF); vRole Physical (RP); Role-emotional (RE); Vitality (VT); Mental Health (MH); Social functioning (SF); bodily pain (BP); general health (GH).

1.        PSQI scores were significantly different between melatonin and placebo treatments (p<0.0001) showing the melatonin treatment group had lower scores.

2.        Sleep latency scores were not significantly different between treatments (p=0.23).

3.        Sleep efficiency scores were significantly different between treatments (p=0.04) showing the melatonin treatment had higher scores.

4.        ESS scores were not significantly different between treatments (p=0.15).

5.        HADS anxiety scores were significantly different between treatments (p=0.0006) showing the melatonin treatment had lower scores.

6.        HADS depression scores were not significantly different between treatments (p=0.68).

7.        FSS scores were significantly different between treatments (p=0.03) showing the melatonin treatment had lower scores.

8.        VT and MH scores of the SF-36 were significantly different between treatments (p=0.03 and p=0.01, respectively) showing the melatonin treatment had higher scores.

9.        The other subscales of the SF-36 were not significantly different between treatments (p>0.05).