Supplementary Online Content

Transcription

1 Supplementry Online Content Breen DP, Vuono R, Nwrthn U, et l. Sleep nd circdin rhythm regultion in erly Prkinson disese. JAMA Neurol. Published online Mrch 31, doi: /jmneurol eappendix. Methods etble 1. Clinicl chrcteristics of intensive sleep cohort versus overll PICNICS cohort etble 2. Clinicl chrcteristics of good versus poor sleepers etble 3. Rest/ctivity (ctigrphy) mesures in PD versus controls etble 4. Polysomnogrphy findings in PD ptients with nd without OSA etble 5. Effect of dopminergic medictions on sleep in PD etble 6. Meltonin prmeters in PD versus controls etble 7. Cortisol prmeters in PD versus controls This supplementry mteril hs been provided by the uthors to give reders dditionl informtion bout their work

2 2 Methods Clinicl ssessments Ptients were recruited to the PICNICS study, community-bsed epidemiologicl study designed to identify ll new cses of Prkinson s disese (PD) in the county of Cmbridgeshire, UK. Ptients were recruited from primry nd secondry cre. In order to ensure tht the cohort ws representtive, ssessments took plce either t the John vn Geest Centre for Brin Repir or the ptient s own home. All ptients met UK Brin Bnk criteri for the dignosis of PD. Ptients completed the Pittsburgh Sleep Qulity Index (PSQI), Epworth Sleepiness Scle (ESS) nd Prkinson s Disese Questionnire (PDQ- 39); nd underwent rnge of other ssessments including the Beck Depression Inventory (BDI), Addenbrooke s Cognitive Exmintion (ACE-R), Apthy Evlution Scle (AES nd AES-compnion), Neuropsychitric Inventory (NPI), Cmbridge Behviourl Inventory (CBI) nd Unified Prkinson s Disese Rting Scle (MDS-UPDRS). Motor phenotype (tremor dominnt (TD) or non-tremor dominnt (non-td)) ws clculted bsed on tremor, git nd posturl instbility subscores from the MDS- UPDRS. 1 Levodop equivlent dily dose (LEDD) ws clculted using the formul proposed by Tomlinson nd collegues 2, nd other concurrent mediction conditions nd tretments were recorded. Following their bseline ssessment, 30 consecutive PICNICS ptients were invited to tke prt in the intensive sleep study (regrdless of whether or not they hd sleep problems). Every PICNICS ptient ws sked to prticipte until the trget recruitment number ws reched. Exclusion criteri were ptients with evidence of n typicl prkinsonin disorder; ptients unble to give informed consent (including those with significnt cognitive impirment); ptients without working knowledge of the English lnguge; nd ptients not ble to sty in hospitl unccompnied overnight. We lso recruited 15 helthy ge nd sexmtched controls vi locl dvertising. Prticipnts who were (or hd been) shift workers, nd PD crers/bed prtners, were not eligible for inclusion in the study. Ptients underwent further subjective sleep ssessment, ctigrphy ssessment, polysomnogrphy nd circdin rhythm nlysis. Actigrphy ssessment These unixil ccelerometers mesure pek intensity of movement ech second, guged by the voltge generted by the ccelerometer, which is then expressed s n ctivity count. The ctiwtches were set to medium sensitivity, with 40 counts defined s wke. Epoch length ws 30 seconds. Ptients were sked to firmly press the indented circle on the upper side of the ctivity monitor to denote lights out nd lights on. When this mrker ws missing or lte, judgement ws mde by the study investigtor, tking into ccount the self-reported bed times (tken from the sleep diry completed during the sme two-week period) nd ctivity level. Dily ctigrphy dt ws downloded, verged nd nlysed using specilist softwre (Actiwtch nd Sleep Anlysis 5.51; Cmbridge Neurotechnology, Cmbridge, UK). Actigrphy ws used to clculte the following nocturnl mesures sleep strt (derived utomticlly from the mrked lights off time), sleep end (derived utomticlly from the mrked lights on time), sleep period (time elpsed from sleep strt to sleep end), totl sleep time (ssumed sleep minus wke time), sleep ltency (time between lights off nd sleep onset), sleep efficiency (percentge of time spent sleep whilst in bed), nd movement nd frgmenttion index (the ddition of percentge of time spent moving nd immobility phses of one minute) which is used s n indictor of restlessness. Non-prmetric circdin rhythm ctivity nlysis ws subsequently crried out using the sme softwre to clculte generl ctivity/rest mesures M10 count (verge ctivity during the most ctive 10 hours), L5 count (verge ctivity during the lst ctive five hours), reltive mplitude (difference between M10 nd L5), intr-dily vribility (quntifies the degree of frgmented motor ctivity during the 24-hour period), inter-dily stbility (quntifies the degree of resemblnce between ctivity ptterns on individul dys), light:drk rtio, verge count during light, nd verge count during drk.

3 3 Polysomnogrphy Ptients were not llowed to np during the dy before the sleep study, nor were they llowed to tke vigorous physicl ctivity or hve cffeinted drinks. Electroencephlogrphy ws crried out using the stndrd interntionl electrode plcement system, together with electrooculogrphy, submentl chin nd bilterl nterior tibilis electromyogrms, n electrocrdiogrm, pulse oximetry, chest nd bdominl movement detector, nsl irflow detector, thermistor nd snoring sensor. The following polysomnogrphy mesures were cptured sleep period time (time elpsed from sleep onset to sleep offset), totl sleep time (ctul sleep time in sleep period), sleep ltency (time between lights out nd sleep onset; defined s three epochs of stge 1 or one epoch of ny other sleep stge), REM ltency (time between sleep onset nd first epoch of REM sleep), proportion of sleep period spent in ech sleep stge, sleep efficiency (rtio of totl sleep time to sleep period time), rousls index (AI) (number of rousls per hour during sleep period), periodic limb movement index (PLMI) (number of sleep-relted periodic limb movements per hour), pnoe-hypopnoe index (AHI) (number of pnoes or hypopnoes per hour), desturtion index (DI) (number of desturtion episodes per hour greter thn 4%) nd minimum nocturnl oxygen sturtion. During the dy, ptients underwent Multiple Sleep Ltency Testing (MSLTs). This consisted of five np opportunities performed t two-hourly intervls, during which the lights were turned off nd the ptient ws sked to relx nd close their eyes. Dytime sleep ltency time (time of first epoch of greter thn 15 seconds of cumultive sleep on 30 second epoch) ws recorded. If the ptient fell sleep, the np ws llowed to continue for further 15 minutes. If no sleep ws observed fter 20 minutes, the np opportunity ended. Men Sleep Ltency (MSL) ws clculted by verging cross ll np opportunities. Dignostic criteri for primry sleep dignoses REM Sleep Behviour Disorder (RBD) ws defined s the presence of REM sleep without toni nd history of injurious, potentilly injurious or disruptive behviour ssocited with sleep nd/or documenttion of bnorml behviours observed on time synchronized infrred video recording during those periods of REM sleep without toni. Significnt Periodic Limb Movements of Sleep (PLMS) ws defined s periodic limb movement index exceeding 15/hour. A dignosis of Restless Legs Syndrome (RLS) relied on the ptient describing the chrcteristic symptoms (typiclly unplesnt senstions in the legs tht occur in the evening nd re relieved by movement). Obstructive Sleep Apnoe (OSA) nd Centrl Sleep Apnoe (CSA) were dignosed ccording to estblished criteri (9), with severity of OSA defined ccording to the pnoe-hypopnoe index: mild (5.1-15/hour), moderte ( /hour) or severe (>30/hour). Excessive Dytime Sleepiness (EDS) ws defined s men sleep ltency (MSL) of less thn eight minutes. Circdin rhythm nlysis Hormone nlysis ELISA ws crried out using the IBL Interntionl GMBH nd R&D Systems ccording to the mnufcturer s instructions. For the quntittive mesurement of meltonin, ech smple ws first pssed through C18 reversed phse column, then extrcted with methnol, evported to dryness nd reconstituted with wter. Following this, ech smple ws dded to the corresponding well of microtiter plte coted with the got-nti-rbbit ntibody. The unknown mount of ntigen present in the smple nd the fixed mount of enzyme-lbelled ntigen competed for ntibody binding sites. After incubtion for one hour t room temperture, the wells were wshed to stop the competition rection. After dding the p-nitrophenyl phosphte substrte solution, the concentrtion of meltonin ws inversely proportionl to the opticl density mesured t 405nm. Opticl densities were mesured using µqun microplte spectrophotometer (Biotek, Winooski, VT, USA). Meltonin stndrds were used to construct clibrtion curve ginst which the concentrtion of unknown smples ws clculted.

4 4 The cortisol ssy ws bsed on the competitive binding technique in which cortisol present in smple competes with fixed mount of horserdish peroxidse-lbelled cortisol for sites on mouse monoclonl ntibody. Following incubtion, the wells were wshed to remove excess conjugte nd unbound smple, nd substrte solution ws dded to the wells to determine the bound enzyme ctivity. The concentrtion of cortisol ws inversely proportionl to the opticl density mesured t 450nm. Bsed on hormone concentrtions t ech time point, we determined the crophse (mximum concentrtion) nd ndir (minimum concentrtion). Amplitude ws defined s hlf the difference between crophse nd ndir. Hormone onset time ws defined s the first sustined rise of two stndrd devitions (SD) bove bseline levels recorded between 13:00-16:00 (meltonin) or 20:30-23:30 (cortisol). Hormone offset time ws defined s the first sustined fll of two stndrd devitions below crophse. We lso clculted the totl 24-hour hormone production (re under curve clculted using the trpezoid rule). Clock gene nlysis Totl cellulr RNA ws isolted from smples of peripherl blood mononucler cells using the PAXgene Blood RNA Kit (Qigen, Hilden, Germny) ccording to the mnufcturer s instructions. On-column DNse tretment ws performed using the RNse-Free DNse Set (Qigen) to void residul genomic DNA contmintion. RNA concentrtions were mesured using the Nnodrop 2000c Spectrophotometer (Thermo Scientific, Wlthm, MA, USA). RNA qulity ws checked by ensuring tht the A260/280 rtio ws greter thn 1.8. The qrt-pcr ws performed on Light Cycler 480 Instrument using the one-step RelTime Redy RNA Virus Mster Kit nd the RelTime Redy Probes (Roche, Mnnheim, Germny). 50ng of RNA per rection nd 10μM of gene specific ssys were used. Cycling prmeters were 8 minutes t 50 C (reverse trnscription), followed by pre-incubtion t 95 C for 30 seconds nd 45 cycles t 95 C for 1 second, 60 C for 20 seconds, nd 72 C for 1 second. The specificity of PCR products ws confirmed using melting curve nlysis. The reltive bundnce of messenger RNA ws clculted using stndrd curve method. The constitutively expressed non-rhythmic -ctin gene ws chosen fter screening pnel of 19 humn housekeeper genes using the RelTime Redy Humn Reference gene pnel. Sttisticl nlysis Dt ws exmined for normlity using visul histogrms nd the Kolmogorov-Smirnov test. Mens were compred using unpired t-tests or ANOVA (normlly distributed dt), nd Mnn-Whitney (non-normlly distributed dt). Chi-squred or Fisher s Exct tests were used to compre proportions between groups. Person or Spermn s rnk correltion coefficients were clculted to ssess bivrite ssocitions. Positive nd negtive predictive vlues were used to evlute the usefulness of screening questionnires for predicting certin primry sleep disorders. Liner regression (continuous dependent vribles) nd logistic regression (ordinl dependent vribles) models were used to control for covrites (sex nd ge unless stted otherwise). Effect sizes were clculted using Glss s delt. Clock gene expression dt ws normlised by clculting z-scores t ech time point. Hormone nd normlised clock gene dt ws then nlysed using repeted-mesures 2-wy ANOVA (djusted for ge nd sex). Muchly s test ws performed to exmine for ny violtions of the ssumption of sphericity, nd, if present, the degrees of freedom were corrected using the Greenhouse-Geiser method. Cosinor nlyses were subsequently pplied to ll normlised circdin dt to define the phse (mesor of the fitted curve) nd rhythmicity (how well the sine curve fitted the dt) of ech individul circdin profile. The threshold of significnce ws set t P<0.05. Sttisticl nlyses were performed using SPSS version 21 (SPSS, Chicgo, IL, USA).

5 5 Online-only References 1. Stebbins GT, Goetz CG, Burn DJ, et l. How to identify tremor dominnt nd posturl instbility/git difficulty groups with the movement disorder society unified Prkinson s disese rting scle: comprison with the unified Prkinson s disese rting scle. Mov Disord 2013;28(5): Tomlinson CL, Stowe R, Ptel S, et l. Systemtic review of levodop dose equivlency reporting in Prkinson's disese. Mov Disord 2010;25(15):

6 6 etble 1. Clinicl chrcteristics of intensive sleep ssessment cohort versus overll PICNICS cohort Vrible PD sleep PICNICS (n=30) (n=239) Mle sex (%) 53% 62% Age t motor symptom onset (yers) b 62 (8) 67 (9) <0.001* Age t dignosis (yers) b 63 (8) 68 (9) 0.009* On dopminergic therpy (%) 80% 42% <0.001* LEDD in treted ptients (mg) 312 (157) 316 (209) MDS-UPDRS prt III c 24 (6) 32 (13) 0.001* Hoehn nd Yhr stge 1.27 (0.45) d 1.80 (0.77) e <0.001* ACE-R 92 (4) 90 (7) Globl PSQI 5.0 (2.9) 6.0 (3.9) ESS 8.4 (4.6) 6.6 (3.8) Results expressed s men (SD) unless stted otherwise *Significnt difference t 0.05 level LEDD=Levodop Equivlent Dily Dose, MDS-UPDRS=Unified Prkinson s Disese Rting Scle, ACE-R=Addenbrooke s Cognitive Exmintion, PSQI=Pittsburg Sleep Qulity Index, ESS=Epworth Sleepiness Scle Chi-squred test; b Unpired t-test; otherwise Mnn Whitney test used; c Bsed on MDS-UPDRS ssessments performed within the lst six months; d Rnge=1-2; e Rnge=1-4

7 7 etble 2. Clinicl chrcteristics of good versus poor sleepers Vrible Domin Good sleepers (n=98) Poor sleepers (n=94) Mle sex (%) Gender Age t dignosis Age 68 (9) 68 (10) b BDI Depression 5.8 (4.3) 8.8 (5.5) <0.001* ACE-R Cognition 91 (6) 88 (6) <0.001* AES Apthy 19 (15) 25 (15) 0.002* AES-compnion c Apthy 18 (15) 25 (17) NPI c Neuropsychitric 2.0 (5.7) 3.6 (5.1) MDS-UPDRS prt I Non motor ADL 6.6 (3.6) 8.9 (3.6) <0.001* MDS-UPDRS prt II Motor ADL 8.4 (5.2) 9.8 (5.1) MDS-UPDRS prt III Motor impirment 29 (11) 33 (12) Hoehn nd Yhr Motor milestones 1.7 (0.7) 1.9 (0.8) TD phenotype (%) Motor phenotype CBI c Behviourl 15 (17) 21 (23) ESS score Dytime sleepiness 6.1 (3.8) 7.3 (3.8) On dopminergic therpy (%) Mediction LEDD (mg) Mediction dose 282 (184) 323 (209) Dopmine gonist (%) Dopmine gonist Results expressed s men (SD) unless stted otherwise *Significnt difference t 0.05 level, fter djustment for multiple comprisons using Bonferonni method BDI=Beck Depression Inventory, ACE-R=Addenbrooke s Cognitive Exmintion, AES=Apthy Evlution Scle, NPI=Neuropsychitric Inventory, MDS-UPDRS=Unified Prkinson s Disese Rting Scle, TD=Tremor dominnt, CBI=Cmbridge Behviourl Inventory, ESS=Epworth Sleepiness Scle, LEDD=Levodop Equivlent Dily Dose Non-completion of questionnires BDI (n=14), AES (n=2), AES-compnion (n=28), NPI (n=52), CBI (n=34) nd ESS (n=22) Chi-squred test; b Unpired t-test; otherwise Mnn-Whitney test used; c Questionnires completed by compnion s instructed

8 8 etble 3: Rest/ctivity (ctigrphy) mesures in PD versus controls Prmeter PD (n=29) Sleep strt 23:36 (00:51) Sleep end 07:05 (00:50) Sleep period 07:28 (00:48) Totl sleep time (mins) 06:40 (00:44) Sleep ltency (mins) 00:04 (00:05) Sleep efficiency (%) 88.5 (4.8) Frgmenttion index 44.4 (20.3) Amplitude (8989) M10 count (9250) L5 count 959 (559) Intr-dily vribility 0.92 (0.16) Inter-dily stbility 0.51 (0.10) Light:drk rtio 3.02 (0.85) Averge during light 113 (71) Averge during drk 39 (22) Controls (n=15) 23:01 (00:53) 06:59 (00:44) 07:57 (00:59) 06:55 (00:40) 00:04 (00:06) 86.9 (5.7) 39.4 (14.9) (6304) (6518) 1145 (693) 0.82 (0.16) 0.54 (0.09) 3.86 (1.03) 145 (51) 39 (14) (univrite) (multivrite) c * b * * b * 0.018* b * 0.002* 0.018* Results expressed s men (SD) unless otherwise stted *Significnt difference t 0.05 level Actigrphy not completed in one PD ptient, therefore this dt excluded from the nlysis; b Unpired t-test; otherwise Mnn- Whitney test used; c Liner regression used, djusting for ge nd sex

9 9 etble 4. Polysomnogrphy findings in PD ptients with nd without OSA Prmeter No OSA (n=14) OSA (n=15) (univrite) (multivrite) c Sleep strt (hrs:min) 22:27 (00:46) 20:38 (05:44) Sleep end (hrs:min) 06:45 (00:29) 07:04 (00:27) Totl sleep time (mins) 381 (50) 432 (76) 0.045* 0.040* Sleep ltency (mins) 12 (13) 15 (14) REM ltency (mins) 111 (61) 122 (128) Sleep efficiency (%) 76 (12) 79 (9) AI (rousls per hour) 19 (6) 23 (7) PLMI (events per hour) 23 (35) 12 (21) AHI (events per hour) 1 (1) 20 (19) <0.001* 0.020* DI (events per hour) 1 (1) 14 (17) 0.010* 0.017* Minimum O 2 sts (%) 90 (4) 85 (4) 0.001* 0.003* MSL (mins) d 12 (5) 10 (5) % time in sleep stges Awke 23 (12) 19 (11) Stge 1 8 (4) 12 (6) Stge 2 49 (10) 49 (9) Stges 3 nd 4 9 (6) 8 (6) REM 11 (6) 13 (7) Results expressed s men (SD) unless otherwise stted All severities (mild, moderte nd severe) included in the OSA group *Significnt difference t 0.05 level AI=Arousls Index, PLMI=Periodic Limb Movement Index, AHI=Apnoe-Hypopnoe Index, DI=Desturtion Index, MSL=Men Sleep Ltency Not possible to distinguish EEG wkefulness from individul sleep stges in one PD ptient, therefore this dt excluded from the nlysis; b Unpired t-test; otherwise Mnn-Whitney test used; c Liner regression used, djusting for ge nd sex (nd, in the cse of AHI, the effect of BMI); d In two PD ptients, MSL bsed on four rther thn five MSLT np opportunities

10 10 etble 5: Effect of dopminergic medictions on sleep in PD Prmeter None (n=10) Dopmine gonist (n=8) b Levodop (n=11) c Both (n=1) d e Totl sleep time (mins) 425 (87) 400 (78) 402 (49) 362 (-) Sleep ltency (mins) 16.8 (18.7) 9.2 (13.5) 12.8 (7.5) 24 (-) REM ltency (mins) 118 (118) 93 (71) 103 (54) 425 (-) Sleep efficiency (%) 79.1 (9.4) 78.8 (12.1) 76.8 (10.5) 64.0 (-) AI (rousls per hour) 22.1 (8.5) 23.3 (6.5) 18.5 (4.9) 23.6 (-) PLMI (events per hour) 24.3 (30.0) 0.3 (0.7) 25.2 (34.6) 0 (-) AHI (events per hour) 6.8 (4.1) 18.4 (24.9) 8.6 (16.7) 19.3 (-) DI (events per hour) 3.5 (3.0) 13.4 (20.7) 5.7 (12.4) 1.2 (-) Minimum O 2 sts (%) 87.2 (4.2) 87.7 (3.1) 87.7 (5.5) 82 (-) MSL (mins) 11.9 (4.7) 7.9 (6.0) 12.5 (3.6) 11.2 (-) % time in sleep stges Awke Stge 1 Stge 2 Stges 3 nd 4 REM 19.0 (11.3) 9.0 (4.9) 48.5 (9.6) 10.4 (8.2) 13.1 (7.6) 21.2 (12.1) 10.9 (6.8) 48.6 (12.3) 7.6 (3.3) 11.7 (6.4) 20.7 (10.8) 10.4 (4.8) 49.2 (8.5) 7.8 (6.0) 11.3 (5.5) 36.0 (-) 8.4 (-) 46.1 (-) 4.3 (-) 5.1 (-) Results expressed s men (SD) AI=Arousls Index, PLMI=Periodic Limb Movement Index, AHI=Apnoe-Hypopnoe Index, DI=Desturtion Index, MSL=Men Sleep Ltency Ptients lso tking rsgiline (n=3) nd mntdine (n=1); b Ptients lso tking rsgiline (n=1) nd mntdine (n=1); c Ptients lso tking rsgiline (n=1) nd mntdine (n=2); d No SD reported becuse only one ptient in this subgroup; e Liner regression used, tking into ccount ny effects of ge, sex nd disese durtion

11 11 etble 6. Meltonin prmeters in PD ptients versus controls Prmeter PD (n=30) Controls (n=15) (univrite) (multivrite) b Acrophse 35.9 (30.3) 43.5 (30.7) Ndir 3.2 (3.0) 6.7 (4.1) 0.005* 0.004* Amplitude 17.0 (15.3) 18.1 (13.1) Onset time c 20:58 (04:23) 17:30 (09:29) Offset time 06:06 (05:11) 05:06 (01:19) Pek durtion c 6.4 (2.4) 5.8 (1.5) Are under curve d (11267) (17497) * Results expressed s men (SD) *Significnt difference t 0.05 level Mnn-Whitney test used; b Liner regression used, djusting for ge nd gender; c Absence of sustined rise in meltonin in 2 PD ptients nd 2 controls, therefore meltonin onset time nd pek durtion could not be determined; d Are under curve clculted using the trpezoid rule (pg/ml*minute)

12 12 etble 7. Cortisol prmeters in PD ptients versus controls Prmeter PD (n=30) Controls (n=15) (univrite) (multivrite) c Acrophse (27.8) 78.4 (31.4) * 0.001* Ndir 8.2 (5.2) 6.0 (4.1) Amplitude 50.6 (13.4) 36.2 (15.4) * 0.002* Onset time d 04:11 (02:28) 04:06 (02:04) Offset time d 08:18 (01:59) 08:43 (01:28) b Pek durtion d 4.6 (2.4) 4.4 (1.8) Are under curve e (16028) (5857) <0.001 * <0.001* Results expressed s men (SD) *Significnt difference t 0.05 level Unpired t-test; b Mnn-Whitney test; c Liner regression used, djusting for ge nd gender; d Pek durtion could not be determined in 7 PD ptients nd 2 controls due to the bsence of sustined cortisol onset (n=3 nd n=0 respectively) or cortisol offset (n=4 nd n=2 respectively); e Are under curve clculted using the trpezoid rule (ng/ml*minute)