Table 2.

Summary of Findings

Oral magnesium supplementation for older adults with insomnia
Population: Older adults ≥55 years old with insomnia Intervention: Oral magnesium supplementation Comparison: Placebo
Outcome (Duration of Follow Up)	No of Particip-ants (Studies)	Absolute Effects (Mean Difference a ± Standard Deviation – unless otherwise specified with *)		Relati-ve Effects b	Quality of Evidence	Vote Count by Direction of Effect	Comments
		Placebo	Magnesium Supplementation
Sleep Parameters
Total sleep time (TST) Time from sleep onset to offset (min) (20 days to 8 weeks)	55 (2)	*The mean TST post-intervention ranged from 326.2 to 456.0 min	*The mean post-intervention TST in the intervention group was 16.06 min higher (95% CI: − 5.99 to 38.12; p = 0.15)	–	Low 1,2	Positive Effect i
Sleep onset latency (SOL) Time from wakefulness to initiation of sleep (min) (20 days to 8 weeks)	55 (2)	*The mean SOL post-intervention ranged from 34.7 to 84.0 min	*The mean post-intervention SOL in the intervention group was − 17.36 min lower (95% CI: − 27.27 to − 7.44, p = 0.0006)	–	Low 1,2	Positive Effect	Lower numbers indicate less night-time wakefulness and better insomnia symptomology of sleep initiation
Sleep efficiency (SE) Sum of REM & non REM sleep / total time in bed (h) (8 weeks)	43 (1)	MD = −  0.00 ± 0.05	MD = −  0.06 ± 0.01 h	–	Low 3	Positive Effect
Early morning awakening (EMA) Premature termination of sleep (h) (8 weeks)	43 (1)	MD = 1.03 ± 0.02	MD = 1.01 ± 0.05	–	Low 3	Null Effect	Lower numbers indicate less early morning awakenings and better insomnia symptomology of sleep maintenance
Slow wave sleep (SWS) NREM stage 3 and 4 sleep (min) (20 days)	12 (1)	MD = + 10.1 ± 15.4	MD = + 16.5 ± 20.4	–	Very Low 1,2,4	Positive Effect	SWS, or deep sleep, is purported to be more restorative sleep.
Sleep Questionnaires
Insomnia Severity Index Score from 0 to 28; ≥ 15 = clinical insomnia (8 weeks)	43 (1)	MD = −  0.5 ± 1.71	MD = −  2.38 ± 2.24	–	Low 3	Positive Effect	Lower scores indicate better sleep quality.
PSQI Score from 0 to 21; ≥ 5 = poor sleeper (8 weeks)	96 (1)	MD = −  4.1 See comment	MD = −  3.4 See comment	–	Low 5	Null Effect ii	No numerical confidence intervals were reported but available in Figure form.
Adverse Events
No data See comment	–	–	–	–	–	–	None of the studies reported adverse events

a – All mean differences (MD) are within group change from baseline mean differences unless otherwise specified with *. The * mean differences are between group post-intervention/treatment mean differences

b – No dichotomized outcomes were reported in any of the studies

Acronyms: h Hour; min Minute; nREM Non rapid eye movement; REM Rapid eye movement

GRADE Working Group grades of evidence

High certainty = very confident that the true effect lies close to that of the estimate of the effect

Moderate certainty = moderately confident that the true effect lies close to that of the estimate of the effect

Low certainty = limited confidence in the effect estimate, the true effect may be substantially different from the estimate of effect

Very Low certainty = very little confidence in the effect estimate, the true effect is likely to be substantially different from the estimate of effect

1 – Serious or concerning methodological limitations were detected in all studies, especially poor internal validity in the randomization process and bias arising from deviations from intended outcomes. Downgrade one level for risk of bias

2 – Only two studies included, with wide confidence intervals and total sample size of 55. Downgrade one level for imprecision

3 – Only one study included. Some concerns for risk of bias in the randomization process and bias arising from deviations from intended outcomes (same as above) mainly due to poor reporting. Downgraded one level. Total sample size of 43. Downgraded one level for imprecision. (Total: 2 levels downgraded)

4 –SWS is a surrogate outcome for insomnia symptoms, the main outcome assessed in the review question. While there is biological plausibility that SWS may help with restorative sleep, there is limited evidence in SWS to improvement in insomnia symptoms. Downgraded one level for indirectness of evidence

5 – Only one study included. High risk of bias from selective reporting. Downgrade one level. Total sample size of 96. Downgraded one level for imprecision

Voting by Direction of Effect

i – Despite lack of statistical significance in the meta-analysis, vote counting was conducted purely by observed direction of effect alone

ii – Each question of the PSQI is scored 0, 1,2 or 3. Thus, a difference of less than 1 is categorized as a null effect

Reference: Schünemann HJ, Higgins JPT, Vist GE, Glasziou P, Akl EA, Skoetz N, Guyatt GH. 2019. Chapter 14: completing ‘summary of findings’ tables and grading the certainty of the evidence. In: Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane handbook for systematic reviews of interventions version 6.0 (updated July 2019). Cochrane. Available from

www.training.cochrane.org/handbook (http://www.training.cochrane.org/handbook)