Table 3.

Characteristics of studies of objective measurement of sleep, HR/HRV and PAL in suicide included for full review.

Study Study design Country	Sample characteristics	Suicide measure(s)	Objective measure(s)	Duration Settings	Key results	STROBE
Dahl et al. (1990) Cross-sectional USA	MDD with attempted suicide or SI (N=13, 54% female, 15.5±1.4 years); MDD without attempted suicide or SI (N=14, 64% female, 15.2±1.5 years); HC (N=30, 57% female, 15.0±1.6 years) Note: Adolescents; Medication not reported	Current SI (the item for SI or attempts on K-SADS-P ≥ 4)	Sleep (EEG): Sleep latency; Awake; Total sleep; Stages 1, 2, 3 and 4; Stage REM; REM period latency; REM density	3 nights Laboratory	↑ Sleep latency in suicidal MDD vs. non-suicidal MDD and HC	20/22
Sabo et al. (1991) Cross-sectional USA	MDD with a history of suicide attempt (N=17, 71% female, 39.3±11.6 years); MDD without a history of suicide attempt (N=17, 71% female, 39.4±12.3 years); HC (N=17, 71% female, 39.3±11.4 years) Note: Inpatients/Outpatients; Free of psychotropic drugs for at least 2 weeks	History of suicide attempt (SADS-L)	Sleep (EEG): Sleep latency; Sleep efficiency; Late-night delta counts; REM time; Number of REM periods; % REM sleep; REM sleep activity; Minimum REM latency; REM density; REM intensity	Overnight Clinic	↑ Sleep latency in attempters vs. HC; ↓ Sleep efficiency and late-night delta counts in attempters vs. HC; ↑ REM time and activity in period 2, and more delta wave counts in NREM period 4 in attempters vs. non-attempters	19/22
Keshavan et al. (1994) Cross-sectional USA	SA/SZ with a history of SI or attempt (N=19, 58% female, 30.6±7.7 years); SA/SZ without a history of SI or attempt (N=20, 40% female, 32.3±8.6 years) Note: Inpatients; Free of psychotropic drugs for at least 2 weeks	History of SI or suicide attempt (review of medical records)	Sleep (EEG): Number of REM periods; REM latency-awake; Total REM time; Total REM activity; Total REM density; REM time, activity and density in REM P1	2 nights Clinic	↑ REM activity in suicidal SA/SZ vs. non-suicidal SA/SZ	18/22
Goetz et al. (2001) Cross-sectional USA	MDD with SI or attempt at baseline (N=18, 14.5±1.3 years); MDD without SI or attempt at baseline but with SI or attempt during approximately 12-year follow-up period (i.e., future suicidality) (N=11, 14.5±2.5 years); MDD with non-SI (N=24, 14.8±1.6 years); HC (N=17, 41% female, 15.1±1.4 years) Note: Adolescents; Gender not reported; Free of psychotropic drugs for at least 2 weeks	Current SI (the item for SI or attempts on K-SADS-P ≥ 4)	Sleep (EEG): Sleep latency; Awake; Total sleep Stages 1, 2, 3 and 4; Stage REM; REM period latency; REM density	3 nights Laboratory	↑ First REM period in MDD with current and future suicidality vs. MDD with non-suicidality and HC	20/22
Agargun and Cartwright. (2003) Cross-sectional USA	MDD with SI (N=13); MDD without SI (N=13) Note: Non-patients depressed sample; All met a diagnosis of current MDD (SCID, BDI ≥ 14 or HDRS ≥ 18); Unmedicated; Gender and age not reported	Current SI (item 9 on BDI ≥ 1 or item 3 on HDRS ≥ 1)	Sleep (EEG): REM latency; % REM sleep; Number of REM periods; Total sleep time; Sleep efficiency	3 nights Laboratory	↓ REM latency in suicidal MDD vs. non-suicidal; ↑ REM % in suicidal MDD vs. non-suicidal	18/22
Lemogne et al. (2011) Cross-sectional France	Individuals who completed suicide (N=133, 20% female, 44.6±13.0 years); Individuals who did not complete suicide (N=323,577, 37% female, 44.7±11.8 years) Note: Medication not reported	Completed suicide (review of medical records); Follow-up time period was 9.1 years on average to death by suicide	HR (ECG): HR	10 minutes Laboratory	↑ HR predicted suicide	21/22
Chang et al. (2012) Cross-sectional Taiwan	MDD with SI (N=261, 49% female, 42.3±14.3 years); MDD without SI (N=237, 52% female, 35.6±13.2 years); HC (N=462, 49% female, 40.7±14.9 years) Note: Outpatients; Free of psychotropic drugs for at least 2 weeks	Current SI (Item 3 on HDRS ≥ 1)	HRV (ECG): HR; SDRR; VLF; LF; HF; LF/HF	5 minutes Laboratory	↑ HR in suicidal MDD vs. non-suicidal MDD; ↓ SDRR, LF and HF in suicidal MDD vs. non-suicidal MDD; ↓ SDRR and HF in suicidal MDD vs. HC	20/22
Chang et al. (2013) Cross-sectional Taiwan	MDD with SI (N=237, 48% female, 40.8±14.3 years); MDD without SI (N=233, 52% female, 38.9±14.1 years); HC (N=462, 49% female, 40.7±14.9 years) Note: Remitted MDD; Outpatients; Free of psychotropic drugs for at least 2 weeks	History of SI during past depression (it was not reported how they assessed)	HRV (ECG): HR; SDRR; VLF; LF; HF; LF/HF	5 minutes Laboratory	↓ SDRR, LF and HF in suicidal MDD vs. non-suicidal MDD and HC	20/22
Singareddy et al. (2013) Cross-sectional USA	Children with self-harm behavior (N=27, 54% female, 8.8±1.9 years); Children without self-harm behavior (N=666, 52% female, 8.8±1.7 years) Note: Self-harm behavior includes suicide attempt; Medication not reported	Current symptoms of self-harm behavior (questions to the parent about self-harm and suicide attempt over the past 2 months)	Sleep (PSG): Sleep latency; REM latency; Total sleep time; Sleep efficiency; % Stages 1 and 2; % REM; Wake after sleep onset; Apnea hypopnea index	Overnight Laboratory	↑ % REM in self-harm behavior group vs. non self-harm behavior group	20/22
Guitierrez et al. (2016) Cross-sectional USA	Veterans with a history of or current SI (N=97, 54.6±11.3 years) Note: Gender and medication not reported	History of SI (review of medical record or SBQ-R ≥ 1)	HRV (ECG): Resting RSA	Duration and setting not reported	No significant association between RSA and SI	15/22
Ballard et al. (2016) Cross-sectional USA	MDD or BPD with SI (N=39, 44% female, 44.0±13.1 years); MDD or BPD without SI (N=26, 69% female, 43.5±12.3 years); Note: Treatment-resistant inpatients; Free of psychotropic drugs for at least 2 weeks (MDD); Lithium or valproate (BPD)	Current SI (item 3 on HDRS ≥ 1 or item 10 on MADRS ≥ 4)	Sleep (EEG): Total sleep time; Wake after sleep onset; Sleep efficiency; Nocturnal wakefulness	Overnight Laboratory	↑ Nocturnal wakefulness between 4am and 4:59am in suicidal group vs. non-suicidal group; Significant association between nocturnal wakefulness between 4am and 4:59am and HDRS and MADRS suicide items	21/22
Forkman et al. (2016) Cross-sectional Germany	Individuals with SI (N=13); Individuals without SI (N=24); Note: Students (N=37, 73% female, 24.5±4.1 years); No history of suicide attempt; Unmedicated	Lifetime SI (Four questions on lifetime SI ≥ 1) (e.g., “I wished I were dead”: 0 “never” to 3 “often”)	HR/HRV (ECG): HR; RMSSD; HF	Duration not reported Setting not reported	Significant correlation between a spectral frequency measures of resting vagally mediated HRV and SI, and mean HR and SI after controlling for depression; A trend correlation between RMSSD and SI	19/22
Wilson et al. (2016) Cross-sectional USA	MDD with a history of suicide attempt (N=13, 100% female, 34.1±10.1 years); MDD without a history of suicide attempt (N=22, 100% female, 31.2±11.5 years) Note: Unmedicated (N=31); Antidepressants (N=4); Borderline personality disorder (N=23); Substance use disorder (N=16)	A history of suicide attempt (it was not reported how the authors confirmed a history of suicide attempt)	HRV (ECG): HF	5 minutes Laboratory	↓ HF during stressor task in attempters vs. non-attempters	21/22
Chang et al. (2016) Cross-sectional Taiwan and Norway	In Taiwan (N=532,932), Individuals who completed suicide (N=569, 42% female, 46.3±15.6 years); Individuals who did not complete suicide (N=532,363, 52% female, 41.4±13.9 years) In Norway (n=74,922), Individuals who completed suicide (N=188, 23% female, 47.9±16.0 years); Individuals who did not complete suicide (N=74,734; 51% female, 49.5±17.6 years) Note: MJ cohort (Taiwan); HUNT cohort (Norway); Medication and diagnosis not reported	Completed suicide (review of medical records)	HR (ECG): HR	Duration not reported Setting not reported	↑ HR in suicidal group vs. in the non-suicidal group in both countries	20/22
Khandoker et al. (2017) Cross-sectional UAE	MDD with SI (N=16, 69% female, 37.5±10.4 years); MDD without SI (N=16, 81% female, 15.2±1.5 years); HC (N=29, 59% female, 28.0±6.4 years) Note: Outpatients; Medication not reported	Current SI (MINI – ‘C’ module > 9)	HR/HRV (ECG): HR; RMSSD; SDRR; VLF; HF; LF; LF/HF	10 minutes Setting not reported	↓ RMSSD in suicidal MDD vs. non-suicidal MDD and HC	19/22
Tsypes et al. (2017) Cross-sectional USA	Individuals with a history of suicide attempt (N=56, 100% female, 35.5±5.4 years); Individuals without a history of suicide attempt (N=56, 100% female, 35.0±6.3 years) Note: Community sample; Medication not reported; Suicide attempt group: Lifetime MDD (N=36), Lifetime anxiety disorder (N=24), Lifetime anorexia or bulimia (N=2), Lifetime alcohol use disorder (N=24), Lifetime substance use disorder (N=21); No suicide attempt group: Lifetime MDD (N=38), Lifetime anxiety disorder (N=16), Lifetime anorexia or bulimia (N=2), Lifetime alcohol use disorder (N=18), Lifetime substance use disorder (N=12)	A history of suicide attempt (One interview question: “A suicide attempt is defined as intentionally hurting yourself with at least some wish to die at that time. How many times have you attempted suicide in your life?”)	HRV (ECG): Resting RSA	2 minutes Laboratory	↓ Resting RSA in attempters vs. non-attempters after controlling for lifetime history of psychiatric diagnoses and current symptoms of depression and anxiety, age, income, race/ethnicity	21/22
Chang et al. (2017) Cross-sectional Taiwan	MDD with current SI (N=58, 40% female, 32.4±11.2 years); MDD without a history of SI (N=61, 39% female, 31.9±12.8 years) Note: Outpatients; Unmedicated	History of SI (C-SSRS)	HR/HRV (ECG): HR; SDRR; VLF; LF; HF; LF/HF	5 minutes Laboratory	↓ SDRR, HF and LF in suicidal MDD vs. non-suicidal MDD after controlling for age, gender, BMI, exercise, smoking, alcohol consumption, stress, depression and cortisol level; Negative correlation between the intensity of SI and SDRR, LF and HF	21/22
Bernert et al. (2017a) Longitudinal USA	Undergraduate students with a history of suicide attempt and recent suicide ideation (≤ 6 months) or no history of suicide attempt but current SI (≤ 1 month) (N=50, 72% female, 19.2±1.4 years) Note: Antidepressants (N=5); SNRI (N=2); Anxiolytics (N=1); Antipsychotic (N=1); Mood stabilizer (N=1)	History of suicide attempt and recent suicide ideation (BSS for Suicide Ideation, Pierce SIS for intent and severity of past suicide attempts)	Sleep (actigraphy): Sleep-onset latency; Total sleep time; Wake after sleep onset; Sleep efficiency; Sleep variability; Time in bed; Nap and sleep interval frequency	3 weeks Home	Significant correlation between total sleep time and sleep variability and changes in BSS for SI between baseline and 7-day follow-up; Significant correlation between sleep variability and changes in BSS for SI between baseline and 21-day follow-up	22/22
Bernert et al. (2017b) Cross-sectional USA	MDD or BPD with current SI (N=31); MDD or BPD without current SI (N=23) Note: Treatment-resistant patients; Free of psychotropic drugs for at least 2 weeks (MDD) Lithium or depakote (BPD); MDD (N=30, 33% female, 47.4±13.3 years); BPD (N=24, 71% female, 44.4±11.6 years)	Current SI (Item 3 on HDRS ≥ 1)	Sleep (EEG): Sleep efficiency; REM latency; Wakefulness after sleep onset; Total sleep time; Percent of REM sleep; Duration of NREM	Overnight Laboratory	↓ Stage 4 NREM sleep in suicidal group vs. non-suicidal group after controlling for depression	21/22
James et al. (2017) Cross-sectional USA	Children with a history of SI and parental criticism (N=20, 35% female, 8.9±1.3 years); Children with a history of SI who did not have critical parents (N=51, 44% female, 8.9±1.3 years); Children with no history of SI but with parental criticism (N=83, 33% female, 9.4±1.5 years); Children with no history of SI nor parental criticism (N=242, 52% female, 9.5±1.5 years); Note: MDD (N=2); GAD (N=12); PTSD (N=6); OCD (N=8); SAD (N=31); Medication not reported	Endorsement in suicide question from K-SADS-P (“sometimes children who get upset or feel bad think about dying or even killing themselves. Has your child ever had these types of thoughts”) Or item 9 on CDI (“I think about killing myself but would not do it” or “I want to kill myself” were coded as having SI”)	HRV (ECG): RSA	12 minutes Laboratory	↔ Changes in RSA between resting and experiencing positive or negative emotions in children with a history of SI and parental criticism; ↓ RSA when experiencing positive or negative emotions compared to resting in all other children	21/22
Littlewood et al. (2019) Longitudinal UK	Individuals experiencing a MDD episode with current SI (N=51, 67% female, 35.5±12.8 years) Note: Antidepressants (N=34)	Current SI (BSS)	Sleep (actigraphy): Sleep efficiency; Sleep onset latency; Total sleep time	1 week Home	Significant correlation between shorter total sleep time and higher next-day suicide ideation	21/22
Boafo et al. (2019) Cross-sectional Canada	MDD with acute risk of suicide (N=17, 82% female, 15.0±1.2 years); HC (N=17, 82% female, 14.9±1.1 years) Note: Adolescents; Inpatients; Antidepressants (N=15); Mood stabilizer (N=1); Antipsychotic (N=4)	Risk of suicide (SBQ-R ≥ 8)	Sleep (PSG): Sleep latency; Total sleep time; REM latency; REM density; Wake after sleep onset; NREM and REM stages	Overnight Laboratory	↑ Sleep latency, REM latency, REM density and relative sleep duration for NREM-stage 1 in suicidal MDD vs. HC	21/22
Benard et al. (2019) Cross-sectional France	BPD with a history of suicide attempt (N=57, 72% female, 47.2±12.6 years); BPD without a history of suicide attempt (N=90, 54% female, 44.8±13.0 years); HC (N=89, 54% female, 39.7±13.4 years) Note: Outpatients; Antidepressants (N=15); Mood stabilizer (N=1); Antipsychotic (N=4)	History of suicide attempt (DIGS or SCID)	Sleep and PAL (actigraphy): Time in bed; Assumed sleep; Actual wake time; Sleep efficiency; Sleep latency; Mean activity score; Fragmentation index; Inter-daily variability; Time of the start of the most active 10-h period in daily activities	3 weeks Home	Suicidal BPD group began their daily activities significantly earlier than non-suicidal BPD and HC groups	22/22

Abbreviations for diagnosis, medications and assessment tools: BDI = Beck depressive inventory, BPD = Bipolar disorder, BSS = Beck scale for suicide ideation, CDI = Children’s depression inventory, C-SSRS = Columbia suicide severity rating scale, DIGS = Diagnostic interview for genetic studies, GAD = Generalized anxiety disorder, HC = Healthy controls, HDRS = Hamilton depression rating scale, K-SADS-P = Kiddie diagnostic schedule for affective disorders and schizophrenia for school age children, MADRS = Montgomery-Asberg depression rating scale, MDD = Major depressive disorder, MINI = Mini-international neuropsychiatric interview, OCD = Obsessive-compulsive disorder, Pierce SIS = Pierce suicidal intent scale, PTSD = Post-traumatic stress disorder, SA/SZ = Schizoaffective disorder/Schizophrenia, SAD = Social anxiety disorder, SADS-L = Schedule for affective disorders and schizophrenia – lifetime version, SBQ-R = Suicide behaviors questionnaire-revised, SCID = Structured clinical interview for DSM-IV axis I disorders, SI = Suicidal ideation, Stimulant/nonstimulant selective serotonin reuptake inhibitors (SNRI).