. 2020 Feb 3;7(3):250–266. doi: 10.1002/mdc3.12899

Table 2.

Clinical trials examining pharmacologic interventions for insomnia in PD

Study	Design	Demographic + PD Staging	Intervention	Treatment Duration	Insomnia/Sleep Assessments	Main Findings
Rios Romenets et al., 201357	Randomized clinical trial	N = 18, 78% men; age 66 ± 12 years; H&Y: 1–3	CBT‐I + daily BLT (n = 6) vs. doxepin 10 mg (n = 6) vs. placebo (PLB; red light; n = 6)	6 weeks	PDSS, PSQI, ISI, SCOPA‐S, sleep diary: B, 6 weeks; PGI‐C, CGI‐C, FSS, ESS	Doxepin improved sleep vs. PLB for: ISI (−9 vs. −2^), SCOPA‐night (−5.2 vs. −2.3^), PGI‐C (1.7 vs. 0.5^), CGI‐C (1.4 vs. 0.3^), FSS (−17.0 vs. 0^)
Stocchi et al, 199846	Double‐blind, crossover clinical trial	N = 40; age 66 ± 20 years, H&Y: 2–4; all participants had motor fluctuations	CDLD CR at bedtime (n = 40) vs. PLB	2 weeks, separated by 10‐day washout	Subjective TST, SOL, awakenings, overall sleep	Sleep measures were not significantly improved; only nocturnal akinesia significantly improved with CDLD vs. PLB
Pahwa et al, 200748	Randomized, double‐blind clinical trial	N = 393; 63% men; age 57 ± 11 years; H&Y: 2.7 ± 0.5	Ropinirole PR (2 mg starting dose, up to 24 mg, n = 202) vs. PLB (n = 191)	24 weeks of PR ropirinole vs. PLB	PDSS	Adjusted treatment difference of 4.7 (95% CI 0.8–8.6^*) in PDSS favoring ropinirole PR vs. PLB
Chaudhuri et al, 201249	Randomized, double‐blind clinical trial (secondary analysis)	N = 182 for PDSS <100; 60% men; age 66 ± 10 years; H&Y: 2–4	Ropinirole PR (2 mg starting dose, up to 24 mg, n = 93) vs. PLB (n = 89)	24 weeks of PR ropirinole vs. PLB	PDSS, PDSS subscales	Δ in PDSS subscales mean treatment difference global sleep quality 3.0 95% CI 0.6–5.5^* and nocturnal motor symptoms 1.7 95% CI 0.5–2.9^**
Xiang et al, 201850	Randomized, double‐blind clinical trial (secondary analysis)	N = 119 for PDSS <90; 50% men; age 61 ± 10 years; H&Y: 2–5	Pramipexole (PPX) IR (n = 60) vs. PPX SR (n = 59)	7‐week titration then 11 weeks of PPX IR or PPX SR stable dose	PDSS, PDSS subscales, ESS	Δ in PDSS subscales Total: SR 28.5* IR 21.7^ Nocturnal motor symptoms: SR 9.6^ IR: 8.9^ Global sleep quality: SR 6.6^** IR: 4.3^* No differences in Δ of the PDSS or its subscales in SR vs. IR
Poewe et al, 200751	Randomized, double blind, clinical trial	N = 505; 63% men; age 64 ± 10 years; H&Y NR	ROT (2–16 mg; n = 204) vs. PPX (up to 4.5 mg daily; n = 201) vs. PLB (n = 101)	7‐week titration then 16 week stable dosing	PDSS	Improvement in PDSS with ROT vs. PLB (4.3 vs. −2.8^) and PPX vs. PLB (4.9 vs. −2.8*)
Trenkwalder et al, 201153	Randomized, double‐blind clinical trial	N = 287; 64% men; age 64 ± 10 years; H&Y: 1–5; all participants had early morning motor symptoms	ROT (2 mg starting dose, up to 16 mg, n = 191) vs. PLB (n = 96)	8‐week titration followed by 4‐week stable dosing	PDSS‐2, individual PDSS‐2 items and subscales, nocturia, NMSS	Greater improvement in PDSS with ROT vs. PLB (LS mean difference − 4.3*), with improvement in 10/15 PDSS items; greater improvement in NMSS with ROT vs. PLB (LS mean difference − 6.7^), including sleep/fatigue (LS mean difference − 2.0^**)
Mizuno et al, 201452	Randomized, double blind, parallel group, clinical trial	N = 414; 41% men; age 66 ± 8 years; H&Y: 2.8 ± 0.6	ROT (2–16 mg; n = 164) vs. ropinirole (up to 15 mg; n = 166) vs. PLB (n = 84)	12‐week titration followed by 4‐week stable dosing	PDSS‐2	Improvement in PDSS‐2 with ROT vs. PLB (LS mean difference ‐2.6***) but equivalent between ROT and ropirinole (−0.7; P = 0.28)
Martinez‐Martin et al, 201554	Prospective, observational study	APO group: N = 43; 49% men; age 62 ± 11 years; H&Y 3 IJLI group: N = 44; 57% men; age 63 ± 9 years; H&Y: 4	APO infusion or IJLI	6 months	NMSS, NMSS domains	Improvements in NMSS total score and sleep/fatigue scale for both treatments Relative change (%) in sleep/fatigue domain was greater for IJLI than APO (−48 vs. −24^*)
Dowling et al, 200558	Double‐blind, crossover clinical trial	N = 40; 73% men; 62 ± 8 years; H&Y: 1.5–5	Melatonin 5 mg (n = 40) vs. melatonin 50 mg (n = 40) vs. PLB (n = 40)	2‐week treatment, separated by 1‐week washout	ESS, SSS, GSDS, PSQI	TST increased with 50 mg melatonin vs. PLB (10 minutes^); GSDS total scores improved with 5 mg melatonin vs. PLB^
Medeiros et al, 200759	Randomized, clinical trial	N = 18; 78% men; 61 ± 7 years; H&Y: 1–3	Melatonin 3 mg (n = 8) vs. PLB (n = 10)	4 weeks	PSG, PSQI, ESS	PSQI scores significant lower with melatonin vs. PLB (4.5 vs. 8.7^*)
Menza et al, 201056	Randomized, double‐blind clinical trial	N = 30; 83% men; mean age 56 years; mean H&Y: 1.6	Eszopiclone (EZP; 3 mg if age < 65; 2 mg if age ≥ 65) vs. PLB	6 weeks	Diary, 10‐point sleep quality scale, PDQ‐8, CGI‐I (sleep), FSS, daytime alertness	Diary showed fewer awakenings with EZP vs. PLB (1.0 vs. 1.8^); CGI‐I for sleep improved more with EZP vs. PLB (2.3 vs. 3.2^); sleep quality improved with EZP vs. PLB (5.0 vs. 5.3^*).
Avila et al, 201560	Prospective observational cohort	N = 24; 50% men; age 75 ± 8 years; median H&Y: 2.0; all with depression	Agomelatine (12.5–50 mg) at bedtime	2‐week titration the 22‐week stable dosing	PDSS, SCOPA‐S, PSG	Relative to baseline, improvements in SCOPA night (6.5 vs. 1.0^), PDSS (95 vs. 119), and PSG awakenings (18 vs. 10^)

Study

Design

Demographic + PD Staging

Intervention

Treatment Duration

Insomnia/Sleep Assessments

Main Findings

Rios Romenets et al., 201357

Randomized clinical trial

N = 18,

78% men;

age 66 ± 12 years;

H&Y: 1–3

CBT‐I + daily BLT (n = 6) vs. doxepin 10 mg (n = 6) vs. placebo (PLB; red light; n = 6)

6 weeks

PDSS, PSQI, ISI, SCOPA‐S, sleep diary:

B, 6 weeks; PGI‐C, CGI‐C, FSS, ESS

Doxepin improved sleep vs. PLB for: ISI (−9 vs. −2^*), SCOPA‐night (−5.2 vs. −2.3^*), PGI‐C (1.7 vs. 0.5^*), CGI‐C (1.4 vs. 0.3^**), FSS (−17.0 vs. 0^*)

Stocchi et al, 199846

Double‐blind, crossover clinical trial

N = 40; age 66 ± 20 years, H&Y: 2–4; all participants had motor fluctuations

CDLD CR at bedtime (n = 40) vs. PLB

2 weeks, separated by 10‐day washout

Subjective TST, SOL, awakenings, overall sleep

Sleep measures were not significantly improved; only nocturnal akinesia significantly improved with CDLD vs. PLB

Pahwa et al, 200748

Randomized, double‐blind clinical trial

N = 393; 63% men; age 57 ± 11 years; H&Y: 2.7 ± 0.5

Ropinirole PR (2 mg starting dose, up to 24 mg, n = 202) vs.

PLB (n = 191)

24 weeks of PR ropirinole vs. PLB

PDSS

Adjusted treatment difference of 4.7 (95% CI 0.8–8.6^*) in PDSS favoring ropinirole PR vs. PLB

Chaudhuri et al, 201249

Randomized, double‐blind clinical trial (secondary analysis)

N = 182 for PDSS <100;

60% men; age 66 ± 10 years;

H&Y: 2–4

Ropinirole PR (2 mg starting dose, up to 24 mg, n = 93) vs. PLB (n = 89)

24 weeks of PR ropirinole vs. PLB

PDSS, PDSS subscales

Δ in PDSS subscales mean treatment difference global sleep quality 3.0 95% CI 0.6–5.5^* and nocturnal motor symptoms 1.7 95% CI 0.5–2.9^**

Xiang et al, 201850

Randomized, double‐blind clinical trial (secondary analysis)

N = 119 for PDSS <90; 50% men; age 61 ± 10 years;

H&Y: 2–5

Pramipexole (PPX) IR (n = 60) vs. PPX SR (n = 59)

7‐week titration then 11 weeks of PPX IR or PPX SR stable dose

PDSS,

PDSS subscales,

ESS

Δ in PDSS subscales

Total:

SR 28.5*** IR 21.7^**

Nocturnal motor symptoms:

SR 9.6^** IR: 8.9^**

Global sleep quality:

SR 6.6^** IR: 4.3^*

No differences in Δ of the PDSS or its subscales in SR vs. IR

Poewe et al, 200751

Randomized, double blind, clinical trial

N = 505; 63% men; age 64 ± 10 years; H&Y NR

ROT (2–16 mg; n = 204) vs. PPX (up to 4.5 mg daily; n = 201) vs. PLB (n = 101)

7‐week titration then 16 week stable dosing

PDSS

Improvement in PDSS with ROT vs. PLB (4.3 vs. −2.8^**) and PPX vs. PLB (4.9 vs. −2.8***)

Trenkwalder et al, 201153

Randomized, double‐blind clinical trial

N = 287; 64% men; age 64 ± 10 years; H&Y: 1–5; all participants had early morning motor symptoms

ROT (2 mg starting dose, up to 16 mg, n = 191) vs. PLB (n = 96)

8‐week titration followed by 4‐week stable dosing

PDSS‐2, individual PDSS‐2 items and subscales, nocturia, NMSS

Greater improvement in PDSS with ROT vs. PLB (LS mean difference − 4.3***), with improvement in 10/15 PDSS items; greater improvement in NMSS with ROT vs. PLB (LS mean difference − 6.7^**), including sleep/fatigue (LS mean difference − 2.0^**)

Mizuno et al, 201452

Randomized, double blind, parallel group, clinical trial

N = 414; 41% men; age 66 ± 8 years; H&Y: 2.8 ± 0.6

ROT (2–16 mg; n = 164) vs. ropinirole (up to 15 mg; n = 166) vs. PLB (n = 84)

12‐week titration followed by 4‐week stable dosing

PDSS‐2

Improvement in PDSS‐2 with ROT vs. PLB (LS mean difference ‐2.6***) but equivalent between ROT and ropirinole (−0.7; P = 0.28)

Martinez‐Martin et al, 201554

Prospective, observational study

APO group:

N = 43; 49% men; age 62 ± 11 years; H&Y 3

IJLI group:

N = 44; 57% men; age 63 ± 9 years; H&Y: 4

APO infusion or IJLI

6 months

NMSS,

NMSS domains

Improvements in NMSS total score and sleep/fatigue scale for both treatments

Relative change (%) in sleep/fatigue domain was greater for IJLI than APO (−48 vs. −24^*)

Dowling et al, 200558

Double‐blind, crossover clinical trial

N = 40; 73% men; 62 ± 8 years; H&Y: 1.5–5

Melatonin 5 mg (n = 40) vs. melatonin 50 mg (n = 40) vs. PLB (n = 40)

2‐week treatment, separated by 1‐week washout

ESS, SSS, GSDS, PSQI

TST increased with 50 mg melatonin vs. PLB (10 minutes^*); GSDS total scores improved with 5 mg melatonin vs. PLB^*

Medeiros et al, 200759

Randomized, clinical trial

N = 18; 78% men; 61 ± 7 years; H&Y: 1–3

Melatonin 3 mg (n = 8) vs. PLB (n = 10)

4 weeks

PSG, PSQI, ESS

PSQI scores significant lower with melatonin vs. PLB (4.5 vs. 8.7^*)

Menza et al, 201056

Randomized, double‐blind clinical trial

N = 30; 83% men; mean age 56 years; mean H&Y: 1.6

Eszopiclone (EZP; 3 mg if age < 65; 2 mg if age ≥ 65) vs. PLB

6 weeks

Diary, 10‐point sleep quality scale, PDQ‐8, CGI‐I (sleep), FSS, daytime alertness

Diary showed fewer awakenings with EZP vs. PLB (1.0 vs. 1.8^*); CGI‐I for sleep improved more with EZP vs. PLB (2.3 vs. 3.2^*); sleep quality improved with EZP vs. PLB (5.0 vs. 5.3^*).

Avila et al, 201560

Prospective observational cohort

N = 24; 50% men; age 75 ± 8 years; median H&Y: 2.0; all with depression

Agomelatine (12.5–50 mg) at bedtime

2‐week titration the 22‐week stable dosing

PDSS, SCOPA‐S, PSG

Relative to baseline, improvements in SCOPA night (6.5 vs. 1.0^**), PDSS (95 vs. 119***), and PSG awakenings (18 vs. 10^*)

P < 0.05;

^**

P < 0.01;

^***

P < 0.001.

PD, Parkinson's disease; H&Y, Hoehn & Yahr; CBT‐I, Cognitive Behavioral Therapy for Insomnia; BLT, bright light therapy; PDSS, Parkinson's Disease Sleep Scale; PSQI, Pittsburgh Sleep Quality Index; ISI, Insomnia Severity Index; SCOPA‐S, Scales for Outcomes in Parkinson's Disease; PGI‐C, patient global impression of change; CGI‐C, clinical global impression of change; ESS, Epworth sleepiness scale; FSS, Fatigue Severity Scale; PDSS‐2, Parkinson's disease sleep scale 2; PLB, placebo; CDLD, carbidopa–levodopa; CR, continuous release; TST, total sleep time; SOL, sleep onset latency; PR, prolonged release; CI, confidence interval; SR, sustained‐release; IR, immediate‐release; ROT, rotigotine; PPX, pramipexole; NMSS, Non‐Motor Sleep Scale; APO, apomorphine; IJLI, intrajejunal levodopa infusion; AE, adverse events; SSS, Stanford Sleepiness Scale; GSDS, General Sleep Disturbance Scale; NR, not reported; PSG, polysomnography; LS, least squares; PDQ‐8, Parkinson's Disease Questionnaire, Short Form; CGI‐I, clinical global impression of improvement; GSDS, General Sleep Disturbance Scale; EZP, eszopiclone.