Table 3.

Meta-analyses of the relationship between sleep and glycemic control in patients with type 1 diabetes (T1D).

Sleep variables	Analysis	Sleep measurements	Studies (n)	N	Type of participants	Resultsa
Sleep stages	MD in percentages of sleep stages between those with optimal and suboptimal glycemic controlsb	PSG, wireless sleep monitor	5	36 vs 81	Adults	No differences in light sleep, MD = −2.90%, (95% CI = −6.96, 1.16).
				36 vs 81		No differences in deep sleep, MD = 2.95%, 95% CI = −1.98, 7.88
Sleep duration	MD in HbA1c levels between those with longer and shorter sleep durationsc	PSG, wireless sleep monitor or actigraphy	6	127 vs 68	Adults	No differences in HbA1c, MD = 0.03%, −0.43, 0.49
		Questionnaire	4	381 vs 152	Adults	Those with longer sleep duration had lower HbA1c, MD −0.24%, 95% CI = −0.47, −0.02.
		Questionnaire	4	96 vs 35	Adolescents/children	No differences in HbA1c in all combined age groups, MD = −0.07%, 95% CI =−0.52, 0.39; age 6–13, MD = 0.07%, 95% CI = −0.42, 0.55; and age >13–17, MD = −0.97%, 95% CI = −2.22, 0.29
	MD in sleep duration between those with optimal and suboptimal glycemic controlsb	PSG, wireless sleep monitor, or actigraphy	6	54 vs 142	Adults	No differences in sleep duration, MD = −2.88 min, 95% CI = −18.09, 12.34
		Questionnaire	4	138 vs 397	Adults	Those with optimal glycemic control had longer sleep duration, MD = 17.28 min, 95% CI = 4.13, 30.370
		Questionnaire	4	35 vs 104	Adolescents/children	No difference in sleep duration in all combined age groups, MD = 18.6 min, 95% CI = −12.6, 49.8; age 6–13, MD = 0.6 min, 95% CI = −39.0, 40.2; and age >13–17, MD = 48 min, 95% CI = −3.0, 99.0
Sleep qualityd	MD in HbA1c levels between those with good and poor sleep qualitye	PSG or actigraphy	4	86 vs 80	Adults	No differences in HbA1c, MD = 0.01%, 95% CI = −0.35, 0.36
		Questionnaire	3	442 vs 136	Adults	Those with good sleep quality had lower HbA1c, MD 95% CI = −0.19%, −0.30, −0.08.
	MD in sleep qualityd between those with optimal and suboptimal glycemic controlb	PSG or actigraphy	5	48 vs 133	Adults	No differences in sleep efficiency, MD = −0.11%, 95% CI = −1.69, 1.47
OSA	MD in HbA1c levels between those with and without OSAf	PSG or oximetry	4	96 vs 81	Adults	No difference in HbA1c, MD = 0.17%, 95% CI = −0.22, 0.57
	MD in HbA1c levels between those with moderate-severe OSA and without OSAf	PSG	3	47 vs 69	Adults	No statistically significant differences in HbA1c, MD = 0.39%, −0.08, 0.87
	MD in AHI between those with optimal and suboptimal glycemic controlsb	PSG	4	53 vs 114	Adults	Those with optimal glycemic control had lower AHI, MD = −2.95 events/h, 95% CI = −5.69, −0.21.

Abbreviations: AHI, apnea–hypopnea index; MD, mean difference; OSA, obstructive sleep apnea; PSG, polysomnography.

^aCalculated by sleep variables of patients with optimal glycemic control minus those of patients with suboptimal glycemic control, or HbA1c of patients with good sleep minus HbA1c of patients with poor sleep, unless otherwise noted.

^bOptimal glycemic control is defined as HbA1c <7% in adults or <7.5% in children, and suboptimal glycemic control is defined as HbA1c ≥7% in adults or ≥7.5% in children, with the exception of the study by Villa et al. [42], in which optimal glycemic control was defined as HbA1c <8%.

^cLonger sleep duration is defined as sleep duration of >6 hours in adults or >8 hours in children, and shorter sleep duration is defined as sleep duration ≤6 hours in adults or ≤8 hours in children.

^dMeasured as sleep efficiency by PSG or actigraphy, or sleep quality score per the sleep questionnaires.

^eGood sleep quality is defined as sleep efficiency ≥85% as measured by PSG or actigraphy or per the cutoff of the sleep questionnaire; poor sleep quality is defined as sleep efficiency <85% as measured by PSG or actigraphy or per the cutoff of the original sleep questionnaire.

^fObstructive sleep apnea (OSA) is defined as AHI ≥5 as measured by PSG or oximetry or having pathological oximetry readings; moderate to severe OSA is defined as AHI ≥15.