Subjective sleep measurement: comparing sleep diary to questionnaire

Figures & data

Table 1 Baseline characteristics of analyzed participants in the retirement and sleep trajectories study (n=1,516)

Variable	Mean (SD) or Percent^a
Any insomnia symptoms^b (%)	57.78
Insomnia days per month^c	21.26 (18.72)
Self-reported sleep apnea (%)	17.22
Age (years)	63.0 (6.89)
Female (%)	53.03
Current smoker (%)	45.84
Body mass index (kilograms/meters^2)	29.48 (6.60)
SF-12 physical health score (points, range 0–100)	48.21 (10.22)
SF-12 mental health score (points, range 0–100)	51.69 (8.72)

Notes: ^aBaseline characteristics are measured using the first of four potential questionnaire-diary pairs received from each participant. ^bFour questionnaire items assessed insomnia symptoms (difficulty falling asleep; difficulty getting back to sleep; repeated nocturnal awakenings; early morning awakening), to which the respondent could report “never,” “rarely,” “sometimes,” “often,” or “almost always.” We considered responses of “often” or “almost always” to any item as reporting any insomnia symptom. ^cRespondents estimated the number of symptomatic days in the last month for each insomnia symptom item (see superscript b). The sum of these symptomatic days is the insomnia days per month.

Abbreviations: SD, standard deviation; SF-12, 12 Item Short Form Health Survey.

Figure 1 Histograms of baseline diary-reported and questionnaire-reported overall sleep durations in the retirement and sleep trajectories study (n=1,516)^a. Key: Dark gray indicates questionnaire-reported sleep duration; light gray indicates sleep diary-reported sleep duration; dashed lines indicate the mean sleep duration for their respective measure. Each bar represents a 15 min block of sleep duration. ^aThere were 5,313 questionnaire-diary pairs for 1,516 study participants, and each participant returned at least one of potentially four questionnaire-diary pairs throughout the study duration. We used baseline questionnaire-diary pairs to generate the histograms.

Table 2 Self-reported sleep duration measures on the sleep diary and questionnaire among participants in the retirement and sleep trajectories study at baseline (n=1,516)^a

Time Frame	Sleep Diary (hours)		Questionnaire (hours)		Difference (hours)^b
	Mean (SD)	Range	Mean (SD)	Range	Mean (SE)	95% CI	P-Value^c
Overall	7.76 (1.01)	4.00, 16.75	7.07 (1.04)	3.20, 11.30	0.69 (0.04)	0.62, 0.76	<0.001
Weekday	7.67 (1.08)	3.25, 16.75	6.94 (1.10)	2.50, 12.00	0.73 (0.04)	0.65, 0.81	<0.001
Weekend	8.03 (1.31)	1.50, 20.00	7.43 (1.16)	2.50, 12.00	0.60 (0.04)	0.51, 0.68	<0.001

Notes: ^aSleep durations were abstracted from the first of potentially four questionnaire-diary pairs received from each participant. ^bDifference = (sleep diary-reported duration) - (questionnaire-reported duration). ^cDifference in mean self-reported sleep durations tested with Welch's two-sample t-tests of difference in means.

Abbreviations: CI, confidence interval; SD, standard deviation; SE standard error.

Figure 2 Bland-Altman plots of diary-reported and questionnaire-reported sleep durations by insomnia symptom status (A) and sleep apnea status in the retirement and sleep trajectories study (B) (n=5,313)^a. Key: Circles indicate “No insomnia symptoms”; x’s indicate “Any insomnia symptom”; squares indicate “No self-reported sleep apnea”; stars indicate “Self-reported sleep apnea”; solid lines indicate the mean sleep duration difference; dashed lines indicate plus or minus two standard deviations from the mean sleep duration difference. Average of Sleep Durations = (diary-reported overall sleep duration + questionnaire-reported overall sleep duration)/2. Sleep Duration Difference = (diary-reported overall sleep duration) – (questionnaire-reported overall sleep duration). ^aThere were 5,313 questionnaire-diary pairs for 1,516 study participants, and each participant returned at least one of potentially four questionnaire-diary pairs throughout the study duration.

Table 3 Demographic predictors of the difference between diary-reported and questionnaire-reported overall sleep duration in the retirement and sleep trajectories study, with dichotomous insomnia exposure (n=1,516)

Variable	Model 1A		Model 1B		Model 1C
	β, hours^a	95% CI	β, hours^a	95% CI	β, hours^a	95% CI
Any insomnia symptom^b
No	Referent		Referent
Yes	0.28***	0.22, 0.33	0.28***	0.23, 0.34
Self-reported sleep apnea
No	Referent				Referent
Yes	0.06	−0.05, 0.16			0.04	−0.06, 0.15
Age (years)	0.00	−0.07, 0.08	0.00	−0.08, 0.08	0.01	−0.07, 0.10
Age^2 (years^2)	0.00	−0.00, 0.00	0.00	−0.00, 0.00	−0.00	−0.00, 0.00
Sex
Male	Referent		Referent		Referent
Female	0.13**	0.05, 0.21	0.13**	0.05, 0.21	0.13**	0.05, 0.21
BMI (kg/m^2)	0.01	−0.02, 0.04	0.01	−0.02, 0.04	0.01	−0.03, 0.04
BMI^2 ((kg/m^2)^2)	−0.00	−0.00, 0.00	−0.00	−0.00, 0.00	−0.00	−0.00, 0.00
Current smoker
No	Referent		Referent		Referent
Yes	0.05	−0.03, 0.13	0.05	−0.03, 0.13	0.05	−0.03, 0.13
SF-12 physical health score (points)	−0.01***	−0.01, −0.00	−0.01***	−0.01, −0.00	−0.01***	−0.01, −0.01
SF-12 mental health score (points)	−0.01***	−0.01, −0.01	−0.01***	−0.01, −0.01	−0.01***	−0.01, −0.01

Notes: *P-value <0.05 **P-value <0.01 ***P-value <0.001. ^aThe beta coefficient is the estimated average incremental difference between diary-reported sleep duration and questionnaire-reported sleep duration in hours per unit change in predictor variables (difference = sleep duration_diary – sleep duration_{questionnaire}). ^bFour questionnaire items assessed insomnia symptoms (difficulty falling asleep; difficulty getting back to sleep; repeated nocturnal awakenings; early morning awakening), to which the respondent could report “never,” “rarely,” “sometimes,” “often,” or “almost always.” We considered responses of “often” or “almost always” to any item as reporting any insomnia symptom.

Abbreviations: β, beta coefficient; BMI, body mass index; CI, profile confidence interval; kg, kilograms; m, meters; SE, standard error.

Table 4 Demographic predictors of the difference between diary-reported and questionnaire-reported overall sleep duration in the retirement and sleep trajectories study, with insomnia days per month exposure (n=1,516)

	Model 2A		Model 2B		Model 2C^b
Variable	β, hours^a	95% CI	β, hours^a	95% CI	β, hours^a	95% CI
Insomnia days per month^c	0.01***	0.01, 0.02	0.01***	0.01, 0.02
Self-reported sleep apnea
No	referent				referent
Yes	0.08	-0.02, 0.18			0.04	-0.06, 0.15
Age (years)	-0.00	-0.08, 0.07	-0.00	-0.08, 0.07	0.01	-0.07, 0.09
Age^2 (years^2)	0.00	-0.00, 0.00	0.00	-0.00, 0.00	-0.00	-0.00, 0.00
Sex
Male	referent		referent		referent
Female	0.12**	0.05, 0.20	0.12**	0.04, 0.20	0.13**	0.05, 0.21
BMI (kg/m^2)	0.01	-0.02, 0.04	0.01	-0.02, 0.04	0.01	-0.03, 0.04
BMI^2 ((kg/m^2)^2)	-0.00	-0.00, 0.00	-0.00	-0.00, 0.00	-0.00	-0.00, 0.00
Current smoker
No	referent		referent		referent
Yes	0.04	-0.03, 0.12	0.05	-0.03, 0.12	0.05	-0.03, 0.13
SF-12 physical health score (points)	-0.00**	-0.01, -0.00	-0.01**	-0.01, -0.00	-0.01***	-0.01, -0.01
SF-12 mental health score (points)	-0.01**	-0.01, -0.00	-0.01**	-0.01, -0.00	-0.01***	-0.01, -0.01

Notes: *P-value <0.05 **P-value <0.01 ***P-value <0.001. ^aThe beta coefficient is the estimated average incremental difference between diary-reported sleep duration and questionnaire-reported sleep duration in hours per unit change in predictor variables (difference = sleep duration_diary – sleep duration_{questionnaire}). ^bModel 2C is equivalent to Model 1C, presented in Table 3. ^cFour questionnaire items assessed insomnia symptoms (difficulty falling asleep; difficulty getting back to sleep; repeated nocturnal awakenings; early morning awakening). Respondents estimated the number of symptomatic days in the last month for each insomnia symptom item. The sum of these symptomatic days is the insomnia days per month.

Abbreviations: β, beta coefficient; BMI, body mass index; CI, profile confidence interval; kg, kilograms; m, meters; SE, standard error.