Enhancements to the multiple sleep latency test

Figures & data

Table 1 Results of sleep questionnaire prior to polysomnography

Question	All (n=30)	SOL <8 minutes (n=13)	SOL >8 minutes (n=17)	P-value
Time to fall asleep (minutes)	22±15	17±15	25±16	NS
Average sleep duration/night (hours)	7.0±1.2	6.9±1.2	7.0±1.1	NS
Number of day naps/week	1.9±2.0	1.0±2.0	2.5±2.5	<0.05
Total sleep/day (hours)	7.7±1.1	7.3±1.1	8.0±1.0	NS
Rested when up in morning	4	3	1	NSa
Strong urges to fall asleep	24	10	14	NSa
More tired than should be	28	11	17	NSa
Driving problems	7	3	4	NSa
Motor vehicle accidents	3	3	0	<0.05a

Notes:

a By Chi-squared test. SOL represents the average SOL during the multiple sleep latency test in minutes.

Abbreviations: SOL, sleep onset latency; NS, not significant.

Table 2 Patient characteristics and results of overnight PSG

Number of patients	30
Sex (F/M)	19/11
Age (years)	37.1±13.6(19–65)
Body mass index (kg/m^2)	27.6±5.3 (19–37)
Idiopathic hypersomnia/no pathology	13/17
Epworth scale	15.2±3 (11–21)
Medications before PSG
No meds	20
Antidepressants	6
Benzodiazepines (clonazepam)	2
Hypnotics (zopiclone)	3
Nonsteroidal anti-inflammatories	6
Opiates (morphine, oxycodone)	2
Total sleep time (minutes)	321±75 (143–443)
Sleep efficiency (%)	75±17 (34–97)
Sleep onset (minutes)	18±15 (1–71)
REM latency (minutes)	189±89 (64–424)
Stage N1 (minutes)	39±16 (15–71)
Stage N2 (minutes)	172±49 (75–301)
Stage N3 (minutes)	65±49 (0–159)
Stage REM (minutes)	44±25 (0–91)
Arousal index (hour^–1)	28±12 (12–54)
Apnea–hypopnea index (hour^–1)	1.8±1.5 (0–5.9)a
Oxygen saturation (%)	96±1 (94–98)
PLM index (hour^–1)	9.5±13 (0–46)b
PLM index with arousal (hour^–1)	2.4±2.8 (0–12)
Odds ratio product
Wake	2.1±0.1 (1.8–2.3)
N1	1.3±0.4 (0.6–1.9)
N2	1.0±0.5 (0.3–1.8)
N3	0.5±0.4 (0.1–1.1)
REM	1.0±0.5 (0.5–2.1)

Notes: Data given as mean ± standard deviation (range).

a Only one patient >5 hour^–1.

b Six patients >15 hour^–1.

Abbreviations: F, female; M, male; PSG, polysomnography; REM, rapid eye movement; PLM, periodic leg movement.

Table 3 Odds ratio product during nocturnal polysomnography in current patients vs patients with mixed sleep disorders

Sleep stage	Mixed disordersa	Current study	SOL <8 minutes	SOL >8 minutes
	n=102	n=30	n=13	n=17
Awake	2.04±0.16	2.11±0.13*	2.11±0.11*	2.10±0.15
non-REM 1	1.04±0.22	1.27±0.33***	1.28±0.40**	1.27±0.30**
non-REM 2	0.68±0.23	0.91±0.42***	0.94±0.43**	0.89±0.43**
non-REM 3	0.36±0.18	0.44±0.33*	0.42±0.29	0.46±0.36*
REM	1.00±0.32	1.05±0.39	1.06±0.50	1.03±0.29

Notes: Results from 102 patients including 49 patients with obstructive sleep apnea (apnea–hypopnea index 26.4±27.1), 14 patients with insomnia, 12 patients with periodic limb movements, and 27 with no pathology. Asterisks, significantly higher than mixed disorders (df >100),

* P<0.05 (t=1.74–1.98),

** P<0.001 (t=3.09–3.65), and

*** P<0.0001 (t=4.02–4.38).

a Republished with permission of American Academy of Sleep Medicine, from Accuracy of automatic polysomnography scoring using frontal electrodes. Younes M, Younes MK, Giannouli E. Vol. 12, number 5, 2016. In press; permission conveyed through Copyright Clearance Center, Inc.

Abbreviations: SOL, average sleep onset latency; REM, rapid eye movement.

Figure 1 Different patterns of evolution of ORP during naps.

Notes: ORP of 2.5 indicates full wakefulness, while an ORP of zero indicates very deep sleep. A–C are different patterns of evolution in ORP. Numbers in brackets are the times of sleep onset identified by the technologist.
Abbreviation: ORP, odds ratio product.

Figure 2 ORP at the time of manually determined sleep onset in 122 naps in which an onset was scored as a function of SOL.

Notes: (A) Individual naps (n=150). (B) Average of five naps in individual patients (n=30). Vertical dashed line in (B) indicates an average SOL of 8 minutes.
Abbreviations: ORP, odds ratio product; NS, not significant; SOL, sleep onset latency.

Figure 3 Examples of ORP patterns in naps with different ranges of SOLs.

Notes: Each line describes the time course of ORP in a representative nap. These examples were chosen at random from all naps having the indicated SOL (method of randomization is described in text).
Abbreviations: ORP, odds ratio product; SOL, sleep onset latency.

Table 4 Frequency and time of reaching different ORP levels (n=150)

Variable	Number reaching target in <20 minutes (naps/patients)	Number reaching target in <10 minutes (naps/patients)	Number reaching target in <5 minutes (naps/patients)
Sleep onset	121/30	89/28	47/21
ORP <2.0	138/30	107/29	77/25
ORP <1.5	112/30	66/27	36/15
ORP <1.0	80/27	31/13	13/9
ORP <0.5	28/15	10/7	2/1

Abbreviation: ORP, odds ratio product.

Figure 4 Relation between average sleep onset latency and the deepest sleep reached during naps.

Notes: Relation between average SOL in individual patients and (A) lowest ORP (ORP_MIN) reached in any of the five naps in the same patient, (B) lowest ORP reached in the first 5 minutes (ORP_MIN(5)) of any of the five naps, (C) average ORP_MIN for the patient, and (D) average ORP_MIN(5) for the patient. Vertical dashed lines indicate an average SOL of 8 minutes.
Abbreviation: ORP, odds ratio product; SOL, sleep onset latency.

Table 5 SOL and ORP variables according to nap order

	SOLa	ORP at SOL	Average ORPb	Lowest ORPa	Time of lowest ORP	Lowest ORP first 5 minutesb
Nap 1	11.5 (7.4)	1.60 (0.34)	1.83 (0.43)	1.12 (0.64)	13.0 (5.5)	1.82 (0.43)
Nap 2	6.3 (5.4)*	1.62 (0.35)	1.52 (0.44)*	0.69 (0.44)*	13.7 (5.1)	1.45 (0.60)*
Nap 3	8.9 (6.8)+	1.50 (0.34)	1.64 (0.44)+	0.80 (0.58)+	13.7 (4.4)	1.68 (0.54)+
Nap 4	9.9 (6.0)	1.63 (0.35)	1.77 (0.39)	0.88 (0.56)	14.8 (4.5)	1.75 (0.46)
Nap 5	11.6 (6.5)	1.75 (0.29)	1.79 (0.36)	1.00 (0.58)	14.6 (5.3)	1.86 (0.47)

Notes:

* Significantly different from all other naps by analysis of variance for repeated measures (df =149;

a F ratio =3.8–4.3;

b F ratio >6.0) and Tukey’s test;

+ Significantly different from naps 1, 2, and 5 (df =149).

Abbreviations: SOL, sleep onset latency; ORP, odds ratio product.

Figure 5 The lowest ORP reached in the first 5 minutes of the nap as a function of nap order.

Notes: Each dot is the average of results in all patients with average SOL >8 minutes (open circles, n=17) and <8 minutes (closed circles, n=13). Bars are SD. ⁺Significantly different from all other naps in patients with SOL >8 minutes (F ratio =3.92, P=0.01). Data points not sharing the same letter are significantly different from each other in patients with SOL <8 minutes (F ratio =5.47, P=0.0008). *t stat >2.43, P=0.01; **t stat =4.38, P=0.0001, significant difference between the same nap in the two groups. Note that the time of most sleepiness is the same in both groups, but patients with SOL <8 minutes are prone to reach deeper sleep in the first 5 minutes at that time.
Abbreviations: ORP, odds ratio product; SOL, sleep onset latency; SD, standard deviation.

Table 6 SOL and ORP variables in naps with and without sleep perception

Variable	With sleep perception	No sleep perception	t-Stat	P
n	12	12
D<2.0	15.4 (2.7)	13.1 (3.3)	3.0	0.006
D<1.5	10.3 (3.6)	8.0 (4.6)	4.1	0.001
D<1.0	6.2 (4.2)	3.8 (3.6)	4.0	0.001
ORPMIN	0.59 (0.35)	0.77 (0.46)	2.7	0.010
Integral ΔORP	37 (7.0)	30 (2.0)	6.1	4.E–05
SOL	6.4 (4.0)	5.7 (2.7)	0.7	0.26

Notes: n denotes the number of patients. ORP_MIN indicates lowest ORP reached. D_<2.0, D_<1.5, and D_<1.0 are the times below indicated ORP values during the nap. Values are means (SD).

Abbreviations: SOL, sleep onset latency; ORP, odds ratio product; SD, standard deviation.

Table 7 SOL and ORP variables in patients with different frequencies of sleep perception

Variable	5/5	4/5	2/5 or 3/5	<2/5	P
n	8	10	7	5
D<2.0	15.1 (4.5)	10.5 (3.3)	12.0 (3.0)	11.4 (4.0)	0.09
D<1.5	10.0 (5.0)	6.4 (2.9)	6.7 (2.6)	5.9 (4.6)	0.16
D<1.0	5.1 (4.6)	3.6 (2.1)	3.4 (2.2)	3.8 (4.3)	0.75
ORPMIN	0.77 (0.33)	0.87 (0.30)	0.97 (0.33)	1.03 (0.56)	0.60
Integral ΔORP	30.8 (8.4)	26.9 (9.6)	24.2 (8.8)	25.7 (17.6)	0.68
SOL	6.2 (3.4)	9.8 (3.7)	11.4 (4.6)	11.4 (5.2)	0.08

Notes: n denotes the number of patients. Column heading shows the number of naps followed by sleep perception. D_<2.0, D_<1.5, and D_<1.0 are the times below indicated ORP values in minutes. ORP_MIN represents the lowest ORP reached. Values are mean (SD).

Abbreviations: SOL, sleep onset latency; ORP, odds ratio product; SD, standard deviation.

Figure 6 Frequency of sleep perception following naps as a function of the decrease in ORP.

Notes: (A) Frequency of sleep perception following naps as a function of the decrease in ORP from its highest value, integrated over the entire nap time in epochs. Open circles are the integral values that are perceived as sleep 50% and 80% of the time, respectively. (B) Average values of perception frequency and ΔORP_INT in individual patients superimposed on the overall relation of Figure 6A.
Abbreviation: ORP, odds ratio product.

Figure 7 Agreement between scorers in scoring sleep onset latency.

Notes: (A) Agreement between the two technologists who scored SOL in the current study. (B) Agreement between the average manual score of SOL and an automatically determined SOL using an ORP <1.6 in two consecutive epochs as a threshold. Note the very high intra-class correlation coefficient in (B) despite the large differences in many naps.
Abbreviations: SOL, sleep onset latency; ORP, odds ratio product; ICC, intra-class correlation.