Comparison of functional disability in older surgical patients with and without probable cognitive impairment: A longitudinal prospective cohort study Authors: Ellene Yan, Yasmin Alhamdah, Nina Butris, Paras Kapoor, Leif Erik Lovblom, Jean Wong, David F. Tang-Wai, Linda Mah, Shabbir MH Alibhai, Maria Carmela Tartaglia, David He, Frances Chung Published: 2026 ABSTRACT BackgroundUnderstanding the recovery trajectory of functional disability in older adults with cognitive impairment can enhance perioperative care through early identification of at-risk patients, risk stratification, and proactive post-discharge planning. This multicenter longitudinal study compared the prevalence and trajectory of functional disability preoperatively and at 30, 90, and 180 days postoperatively between 307 older non-cardiac surgical participants with and without probable cognitive impairment, and examined their incidence of adverse postoperative outcomes.MethodsOnline and telephone assessments were administered to assess cognition and functional disability. Cognition was assessed using the Ascertain Dementia Eight-item Questionnaire (AD8), with a preoperative score ≥2 classifying probable cognitive impairment. Functional disability was self-reported preoperatively and at 30, 90, and 180 days postoperatively using the 12-item World Health Organization Disability Assessment Schedule 2.0, with scores ≥ 16% indicating functional disability exceeding a patient-acceptable symptom state.ResultsOf 307 participants, 17% screened positive for probable cognitive impairment using the AD8 preoperatively. Participants with probable cognitive impairment had a significantly higher perioperative prevalence of functional disability than those without. Nevertheless, probable cognitive impairment did not modify the recovery pattern of functional disability. Additionally, probable cognitive impairment was associated with greater postoperative delirium (OR: 7.37 [95% CI 2.25, 25.97], P = 0.001) and 180-day emergency department visits (OR: 5.05 [95% CI 1.34, 18.92], P = 0.014) after age adjustment.ConclusionsA positive AD8 screen was associated with greater perioperative functional disability and adverse clinical outcomes. Despite this heightened risk, we did not find evidence that probable cognitive impairment modified the recovery trajectory of functional disability. Our exploratory study suggests that the AD8 may be useful in guiding patient-centered preoperative discussions and identifying patients with probable cognitive impairment who may warrant closer perioperative monitoring. Introduction Seventeen percent, or 1.6 billion, of the global population will reach 65 years and older by 2050.1 Given that older adults account for disproportionately higher rates of surgery and adverse outcomes such as complications and mortality, early identification of at-risk patients can guide perioperative care to optimize their recovery.2, 3 Cognitive impairment (CI) is a risk factor for adverse outcomes including delirium, non-home discharge, and one-year mortality.2 The severity of CI ranges from subjective cognitive decline—characterized by cognitive complaints despite performing within the normal range on objective assessments—to mild cognitive impairment (MCI), or mild neurocognitive disorders, and finally to dementia or major neurocognitive disorders. Severe cognitive deficits that can compromise daily functioning characterize dementia. Nearly 37% of older surgical patients with CI remain under-recognized, highlighting an area of improvement in perioperative care.4, 5, 6 In addition to CI, older adults may present other vulnerabilities that can further complicate surgical recovery, such as frailty and functional disability. Both markers of frailty and disability have been associated with adverse postoperative outcomes, with disability serving as the strongest predictor of mortality.7 Specifically, functional disability, defined as the presence of bodily impairment, activity limitations, or participation restrictions in major life areas, is a meaningful patient-reported outcome for surgical populations.8, 9, 10, 11 Given that functional disability imposes significant burden on the quality of life and well-being of older adults, its recovery pattern after surgery requires further characterization. Although previous literature has suggested CI as a risk factor for mortality or functional disability six months after surgery, cognitive screening tools that can be rapidly and remotely administered in the perioperative setting remain underexplored.12 Investigating CI and its associated surgical burden—including patient-reported functional disability—through remote assessment tools provides a practical and novel approach for identifying individuals who may benefit from additional support. Surgical outcomes may be optimized by informing resource allocation for at-risk patients with probable CI and tailoring perioperative care to address their unique needs. While previous publications from this cohort have reported the prevalence of probable CI and the recovery patterns associated with preoperative disability status, the perioperative trajectory of functional disability in participants with and without probable CI has not been characterized previously.13, 14 To address this gap, the aims of this report were to compare (1) the prevalence and trajectory of functional disability preoperatively and at 30, 90, and 180 days postoperatively and (2) the incidence of adverse clinical outcomes between older participants with and without probable CI, using rapid remote cognitive screening tools. We hypothesized that older participants with probable CI would have a higher prevalence of functional disability preoperatively and at 30, 90, and 180 days postoperatively than those without probable CI. Materials and methods This analysis was a component of a multicenter, longitudinal prospective cohort study—the Postoperative Functional Disability in Unrecognized Cognitive Impairment (POF) study.13, 14 The study was approved by the Research Ethics Boards at the University Health Network and Mount Sinai Hospital (REB #: 20–6186.3 and 21–0123E), and written informed consent was obtained from all participating subjects. The trial was registered prior to patient enrollment at www.clinicaltrials.gov (NCT04850833, Date of registration: April 7, 2021). Details of the study have been reported.13, 14 The study was conducted at Toronto Western Hospital and Mount Sinai Hospital, Toronto, Canada from July 23, 2021, to December 5, 2023. Participants aged ≥ 65 years undergoing elective non-cardiac in-patient surgery were recruited consecutively from the preoperative clinics. We excluded participants with diagnosed major neurocognitive disorder, uncorrected visual and/or hearing impairment, current alcohol or substance abuse, or those undergoing ambulatory, vascular, or neurosurgical procedures. Study procedures and measures One to 30 days before surgery, eligible participants provided informed consent and completed the baseline preoperative assessment remotely via an online REDCap survey (www.project-redcap.org) and a telephone call.15 The order in which the online survey and telephone call were administered varied depending on participants’ availability and preferences. Both the online survey and telephone assessments were repeated at 30, 90, and 180 days postoperatively. Participants were also given the option to complete the online survey over the telephone to avoid selecting more computer literate participants. Assessors were blinded to the outcome data to minimize bias. Primary exposure of interest: Probable CI To date, no rapid cognitive screening tools have been validated in surgical populations.16 Additionally, the study protocol was developed during the COVID−19 pandemic with restricted in-hospital visits. Thus, cognitive screening tools were administered either virtually or over the telephone before surgery by trained research personnel to examine their associations with patient-reported outcome measures (PROMs) and adverse clinical outcomes. The primary cognitive screening tool is the Ascertain Dementia Eight-item Questionnaire (AD8),17 which participants completed in an online REDCap survey (www.project-redcap.org).15 Notably, the self-reported AD8 demonstrated a superior diagnostic accuracy for identifying dementia (82% sensitivity and 75% specificity) over MCI (57% sensitivity and 71% specificity).18 In addition to the AD8, the Telephone Montreal Cognitive Assessment (T-MoCA)19, 20 was administered, which has shown 86% sensitivity and 61% specificity for detecting MCI.21 As the AD8 was designed for screening dementia, we further explored a lower cut-off of ≤ 16 on the T-MoCA, which demonstrated a 90% sensitivity and 77% specificity for detecting CI (dementia or MCI) in community and clinic settings.22 The T-MoCA was added after the first 29 participants had been recruited, with one point added for those with ≤ 12 years of education. Their characteristics are presented in supplemental data (Table S1) and described previously.13 In this study, the classification of probable CI and No-CI was based on the AD8 and T-MoCA rather than clinical diagnosis. Tool-specific labels, namely CIAD8/No-CIAD8 and CIT-MoCA/No-CIT-MoCA, denote probable CI as classified by the respective screening instrument. Primary outcome: Functional disability The 12-item World Health Organization Disability Assessment Schedule (WHODAS) 2.0 was adopted as a self-reported measure of functional disability, which has been validated in surgical populations.8, 9, 23 The WHODAS was administered online preoperatively and at 30, 90, and 180 days postoperatively. It contains 12 items scored from zero to four, which evaluates the level of functioning across (1) cognition, (2) self-care, (3) mobility, (4) getting along with others, (5) life activities, and (6) participation in society due to health conditions.9 The score can be converted to a percentage, with a higher score indicating greater functional disability. Specifically, a WHODAS score ≥ 16% exceeds the patient-acceptable symptom state for functional disability, indicating patients may no longer perceive themselves to be well (Table S1).23, 24 A score of ≥ 35% indicates clinically significant or at least moderately severe functional disability. Lastly, a ≥ 5% change in postoperative WHODAS scores from preoperative assessments has been reported as the minimal clinically important difference in surgical patients.23 Given that some of the WHODAS items outside the original cognition domain may also capture aspects of cognitive functioning, the 12 WHODAS items were stratified into cognition-related and non-cognition-related items. Cognition-related items included both items from the original WHODAS domains of cognition, life activities, and getting along with others, as well as one item from participation in society. Non-cognition-related items included both items from the domains of mobility and self-care, and the remaining item from participation in society. Secondary PROMs and adverse clinical outcomes Additional PROMs were assessed in the online survey, including instrumental activities of daily living (Lawton-Brody Instrumental Activities of Daily Living [IADL] scale),25 health-related quality of life (EuroQol 5-dimension Questionnaire [EQ−5D-5L]),26 pain level (visual analog scale [VAS]),27 frailty (5-item FRAIL Scale),28 depression (15-item Geriatric Depression Scale [GDS]),29 and sleep disturbances and quality (Pittsburgh Sleep Quality Index [PSQI] and Single-item Sleep Quality Scale [SQS]).30, 31 The risk of obstructive sleep apnea (OSA) was assessed using the STOP (Snoring, Tiredness, Observed apnea, and High blood pressure) questionnaire.32 Clinical outcomes included postoperative complications (e.g., postoperative delirium and cardiac, respiratory, gastrointestinal, renal, and surgical complications), length of hospital stay (LOS), discharge destination, readmissions, emergency department (ED) visits, reoperation, all-cause mortality, and composite adverse outcomes. Postoperative delirium was evaluated using the Chart-based Delirium Identification Instrument (CHART-DEL) and the short form of Confusion Assessment Method (CAM).33, 34, 35 The CAM was administered every 12 h by the nursing team and daily by trained research assistants from one to three days postoperatively or until hospital discharge. Clinical outcomes were extracted from electronic medical records (EMR) by research personnel blinded to the cognitive assessment outcomes. Additionally, direct questioning about adverse outcomes such as complications was conducted with participants and/or their informants at each follow-up time point over the telephone to supplement EMR. Statistical analysis Data are presented as means ± standard deviations (SD) or standard errors (SE), medians [interquartile ranges (IQR)], or counts (percentages). Unadjusted univariable analyses compared the probable CI and No-CI groups using Student’s t-test, Mann-Whitney U test, Fisher’s exact test, or chi-square test. To examine the binary outcome of functional disability, a generalized estimating equations (GEE) technique was adopted using the dichotomous covariate of probable CI status on the AD8 and T-MoCA, categorical covariate of time, and their interaction; a working independence correlation structure was assumed. The binary outcome of functional disability was analyzed using a cut-off of ≥ 16% (for functional disability) and 35% (for clinically significant functional disability). Post hoc analyses were performed for pairwise comparisons of time using the Tukey method to adjust for multiple comparisons, and adjusted P-values are presented.36 Additional cognitive screening tools, specifically the Centers for Disease Control and Prevention (CDC) cognitive question37 and the Modified Telephone Interview for Cognitive Status (TICS-M),38, 39 were administered but not analyzed due to the limited number of participants screening positive for probable CI. A mixed-effects logistic regression model was also conducted as a sensitivity analysis to address the missing-completely-at-random assumption made by the GEE model. To adjust for confounding, sensitivity analyses included the following clinically relevant variables as additional covariates in the models: age, sex, education level, and American Society of Anesthesiologists (ASA) physical status. Subgroup analyses restricted to orthopedic or spinal surgeries performed with the primary goal of functional improvement were conducted using a similar approach. To examine the continuous outcome of functional disability at all four time points, we used a linear mixed-effects model structure (instead of a GEE), using the same covariate structure and a random participant-level intercept. Model diagnostics were used to verify assumptions, and missing data were accounted for under the Missing-at-random (MAR) assumption within the model, which was considered reasonable. The associations of probable CI with individual WHODAS domains and cognition- and non-cognition-related items were examined using a similar approach. Additionally, linear mixed-effect models were used to explore the association between preoperative AD8 and T-MoCA scores and functional disability across time. Preoperative AD8 and T-MoCA scores were included as continuous covariates in separate models, with their preoperative values repeated for each time point. To explore the associations between probable CI on the AD8 and T-MoCA and adverse clinical outcomes, univariable and multivariable logistic or linear regression analyses were conducted. To address skewness, LOS was log-transformed after adding a value of one. Covariate selection in the multivariable regression models was guided by both clinical relevance and the event counts to avoid overfitting. Specifically, multivariable regression analyses were not conducted for mortality, 30-day readmission, and reoperation due to low event counts. For postoperative delirium and 30-day ED visits, multivariable analyses were minimally adjusted for age due to limited event counts and concern for model overfitting, whereas the remaining 30-day outcomes were fully adjusted for age, sex, and ASA class. For 90-day outcomes of all-cause complications and composite adverse outcomes, multivariable analyses were adjusted for age and sex. The remaining 90-day outcomes and 180-day outcomes were minimally adjusted for age. Given the exploratory and hypothesis-generating nature of these analyses, multiple comparisons were not adjusted to minimize the risk of type II errors, particularly given the limited event counts for several outcomes. Model estimates were expressed as odds ratios (ORs) or beta-coefficients (β) with their 95% confidence intervals (95% CI). All analyses were performed using R (version 4.1.1). Statistical hypothesis tests were two-tailed with an α-level of 0.05. Results Study population We screened 699 participants at Toronto Western Hospital and Mount Sinai Hospital. Three hundred and seven participants consented, with 278 completing all cognitive screening tools preoperatively. After surgery, 216 (70%), 187 (61%), and 183 (60%) participants completed 30-, 90-, and 180-day follow-up assessments, respectively (Fig. S1). Clinical outcomes were collected for all study participants, except postoperative delirium which was missing in eight participants due to conversion to day surgery or early hospital discharge. Seventeen percent of participants screened positive for probable CI on the AD8 (CIAD8), while 28% screened positive on the T-MoCA (CIT-MoCA). The CIAD8 group was significantly more likely to exhibit functional disability (WHODAS ≥ 16%) and clinically significant functional disability (WHODAS ≥ 35%) preoperatively, compared with the No-CIAD8 group (Table 1). They also had poorer IADL, higher pain level, and lower health-related quality of life, as well as a higher prevalence of frailty, depression, and sleep disturbances before surgery (Table 2, Table S2).Table 1Demographics and characteristics of the study population stratified by preoperative probable CI status on the AD8 and T-MoCA.Baseline CharacteristicsAD8T-MoCANo-CIAD8(n = 255)CIAD8(n = 52)No-CIT-MoCA(n = 201)CIT-MoCA(n = 77)Age, years72.0 [68.0–77.0]72.5 [69.0–77.0]72.0 [68.0–76.0]73.0 [69.0–77.0]Sex, female143 (56.1)29 (55.8)118 (58.7)37 (48.1)Education, years16.0 [14.0–16.0]14.5 [12.0–16.5]*16.0 [14.0–18.0]15.0 [12.0–16.0]*Ethnicity/Race White204 (80.0)42 (80.8)163 (81.1)58 (75.3) Asian15 (5.9)1 (1.9)7 (3.5)5 (6.5) Hispanic5 (2.0)2 (3.8)5 (2.5)2 (2.6) Othera10 (3.9)4 (7.7)7 (3.5)7 (9.1) Preferred not to answer21 (8.2)3 (5.8)19 (9.5)5 (6.5)BMI, kg/m229.3 ± 6.731.4 ± 6.5*29.3 ± 6.931.2 ± 6.3*Comorbidities Hypertension129 (50.6)36 (69.2)*102 (50.7)48 (62.3) CAD14 (5.5)4 (7.7)8 (4.0)10 (13.0) Diabetes Mellitus34 (13.3)6 (11.5)23 (11.4)12 (15.6) Asthma/COPD41 (16.1)12 (23.1)33 (16.4)14 (18.2) OSA45 (17.6)18 (34.6)*37 (18.4)21 (27.3) Stroke6 (2.4)3 (5.8)5 (2.5)4 (5.2) Depression16 (6.3)15 (28.8)**18 (9.0)13 (16.9) Cancer74 (29.0)13 (25.0)63 (31.3)19 (24.7) Smoking17 (6.7)3 (5.8)14 (7.0)3 (3.9) Alcohol consumption (Yes vs. No)105 (41.2)19 (36.5)90 (44.8)33 (42.9)ASA Class I2 (0.8)0 (0)2 (1.0)0 (0) II51 (20.0)5 (9.6)26 (17.9)10 (13.0) III187 (73.3)41 (78.8)151 (75.1)60 (77.9) IV15 (5.9)6 (11.5)12 (6.0)7 (9.1)Surgical Procedure Orthopedic167 (65.5)35 (67.3)124 (61.7)55 (71.4) General35 (13.7)4 (7.7)30 (14.9)8 (14.0) Urological18 (7.1)6 (11.5)16 (8.0)5 (6.5) Spinal18 (7.1)1 (1.9)15 (7.5)2 (2.6) Otolaryngological8 (3.1)3 (5.8)6 (3.0)5 (6.5) Otherb9 (3.5)3 (5.8)10 (5.0)2 (2.6)Functional disability Functional disability, ≥16%93 (36.5)40 (76.9)**79 (39.3)39 (50.6) Significant functional disability, ≥35%21 (8.2)21 (40.4)**24 (11.9)14 (18.2)Values expressed as mean ± SD, median [IQR], or count (%) where appropriate. Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; ASA, American Society of Anesthesiologists; BMI, Body Mass Index; CAD, Coronary Artery Disease; CI, probable cognitive impairment as classified by the respective screening instrument; COPD, Chronic Obstructive Pulmonary Disease; OSA, obstructive sleep apnea; T-MoCA, Telephone Montreal Cognitive Assessment. aOther ethnicity/race included Mediterranean, black, Middle Eastern, Indigenous, and mixed. bOther surgical procedures included gynecological, bariatric, and thoracic procedures. ** and * represent P ≤ 0.001 and P ≤ 0.05 in participants with probable CI versus No-CI. A subset of data in this table (demographic and surgical procedures) has been previously published. Reproduced and adapted from Yan et al.13Table 2Preoperative PROMs in participants with versus without probable CI on the AD8.Preoperative PROMsNo-CIAD8(n = 255)CIAD8(n = 52)IADLa7.4 ± 0.16.3 ± 0.2**VAS painb4.2 ± 0.25.5 ± 0.4**EQ−5D-5Lc0.723 ± 0.0120.531 ± 0.026**Frail,≥325 (9.8)13 (25.0)*GDS,≥537 (14.5)25 (48.1)**PSQI, > 5101 (48.6)34 (77.3)**SQSd6.7 ± 0.16.2 ± 0.3Values are expressed as mean ± SE or n (%) where appropriate. Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; CIAD8, probable cognitive impairment as classified by the AD8; EQ−5D-5L, EuroQol 5-dimension Questionnaire; GDS, Geriatric Depression Scale; IADL, Instrumental Activities of Daily Living Scale; PROMs, patient-reported outcome measures; PSQI, Pittsburgh Sleep Quality Index; SQS, Single-item Sleep Quality Scale; VAS, visual analog scale. aIADL scores range from 0 to 8, with a lower score indicating greater dependency in functioning. bVAS pain scores range from 0 to 10, with a higher score indicating a higher pain level. c EQ−5D-5L scores range from −0.148–0.949, with a lower score indicating poorer health-related quality of life. dSQS scores range from 0 to 10, with a lower score indicating poor sleep quality. ** and * represent P ≤ 0.001 and P ≤ 0.05 in participants with CIAD8 versus No-CIAD8. Table 1Demographics and characteristics of the study population stratified by preoperative probable CI status on the AD8 and T-MoCA.Baseline CharacteristicsAD8T-MoCANo-CIAD8(n = 255)CIAD8(n = 52)No-CIT-MoCA(n = 201)CIT-MoCA(n = 77)Age, years72.0 [68.0–77.0]72.5 [69.0–77.0]72.0 [68.0–76.0]73.0 [69.0–77.0]Sex, female143 (56.1)29 (55.8)118 (58.7)37 (48.1)Education, years16.0 [14.0–16.0]14.5 [12.0–16.5]*16.0 [14.0–18.0]15.0 [12.0–16.0]*Ethnicity/Race White204 (80.0)42 (80.8)163 (81.1)58 (75.3) Asian15 (5.9)1 (1.9)7 (3.5)5 (6.5) Hispanic5 (2.0)2 (3.8)5 (2.5)2 (2.6) Othera10 (3.9)4 (7.7)7 (3.5)7 (9.1) Preferred not to answer21 (8.2)3 (5.8)19 (9.5)5 (6.5)BMI, kg/m229.3 ± 6.731.4 ± 6.5*29.3 ± 6.931.2 ± 6.3*Comorbidities Hypertension129 (50.6)36 (69.2)*102 (50.7)48 (62.3) CAD14 (5.5)4 (7.7)8 (4.0)10 (13.0) Diabetes Mellitus34 (13.3)6 (11.5)23 (11.4)12 (15.6) Asthma/COPD41 (16.1)12 (23.1)33 (16.4)14 (18.2) OSA45 (17.6)18 (34.6)*37 (18.4)21 (27.3) Stroke6 (2.4)3 (5.8)5 (2.5)4 (5.2) Depression16 (6.3)15 (28.8)**18 (9.0)13 (16.9) Cancer74 (29.0)13 (25.0)63 (31.3)19 (24.7) Smoking17 (6.7)3 (5.8)14 (7.0)3 (3.9) Alcohol consumption (Yes vs. No)105 (41.2)19 (36.5)90 (44.8)33 (42.9)ASA Class I2 (0.8)0 (0)2 (1.0)0 (0) II51 (20.0)5 (9.6)26 (17.9)10 (13.0) III187 (73.3)41 (78.8)151 (75.1)60 (77.9) IV15 (5.9)6 (11.5)12 (6.0)7 (9.1)Surgical Procedure Orthopedic167 (65.5)35 (67.3)124 (61.7)55 (71.4) General35 (13.7)4 (7.7)30 (14.9)8 (14.0) Urological18 (7.1)6 (11.5)16 (8.0)5 (6.5) Spinal18 (7.1)1 (1.9)15 (7.5)2 (2.6) Otolaryngological8 (3.1)3 (5.8)6 (3.0)5 (6.5) Otherb9 (3.5)3 (5.8)10 (5.0)2 (2.6)Functional disability Functional disability, ≥16%93 (36.5)40 (76.9)**79 (39.3)39 (50.6) Significant functional disability, ≥35%21 (8.2)21 (40.4)**24 (11.9)14 (18.2)Values expressed as mean ± SD, median [IQR], or count (%) where appropriate. Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; ASA, American Society of Anesthesiologists; BMI, Body Mass Index; CAD, Coronary Artery Disease; CI, probable cognitive impairment as classified by the respective screening instrument; COPD, Chronic Obstructive Pulmonary Disease; OSA, obstructive sleep apnea; T-MoCA, Telephone Montreal Cognitive Assessment. aOther ethnicity/race included Mediterranean, black, Middle Eastern, Indigenous, and mixed. bOther surgical procedures included gynecological, bariatric, and thoracic procedures. ** and * represent P ≤ 0.001 and P ≤ 0.05 in participants with probable CI versus No-CI. A subset of data in this table (demographic and surgical procedures) has been previously published. Reproduced and adapted from Yan et al.13 Demographics and characteristics of the study population stratified by preoperative probable CI status on the AD8 and T-MoCA. Values expressed as mean ± SD, median [IQR], or count (%) where appropriate. Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; ASA, American Society of Anesthesiologists; BMI, Body Mass Index; CAD, Coronary Artery Disease; CI, probable cognitive impairment as classified by the respective screening instrument; COPD, Chronic Obstructive Pulmonary Disease; OSA, obstructive sleep apnea; T-MoCA, Telephone Montreal Cognitive Assessment. aOther ethnicity/race included Mediterranean, black, Middle Eastern, Indigenous, and mixed. bOther surgical procedures included gynecological, bariatric, and thoracic procedures. ** and * represent P ≤ 0.001 and P ≤ 0.05 in participants with probable CI versus No-CI. A subset of data in this table (demographic and surgical procedures) has been previously published. Reproduced and adapted from Yan et al.13 Table 2Preoperative PROMs in participants with versus without probable CI on the AD8.Preoperative PROMsNo-CIAD8(n = 255)CIAD8(n = 52)IADLa7.4 ± 0.16.3 ± 0.2**VAS painb4.2 ± 0.25.5 ± 0.4**EQ−5D-5Lc0.723 ± 0.0120.531 ± 0.026**Frail,≥325 (9.8)13 (25.0)*GDS,≥537 (14.5)25 (48.1)**PSQI, > 5101 (48.6)34 (77.3)**SQSd6.7 ± 0.16.2 ± 0.3Values are expressed as mean ± SE or n (%) where appropriate. Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; CIAD8, probable cognitive impairment as classified by the AD8; EQ−5D-5L, EuroQol 5-dimension Questionnaire; GDS, Geriatric Depression Scale; IADL, Instrumental Activities of Daily Living Scale; PROMs, patient-reported outcome measures; PSQI, Pittsburgh Sleep Quality Index; SQS, Single-item Sleep Quality Scale; VAS, visual analog scale. aIADL scores range from 0 to 8, with a lower score indicating greater dependency in functioning. bVAS pain scores range from 0 to 10, with a higher score indicating a higher pain level. c EQ−5D-5L scores range from −0.148–0.949, with a lower score indicating poorer health-related quality of life. dSQS scores range from 0 to 10, with a lower score indicating poor sleep quality. ** and * represent P ≤ 0.001 and P ≤ 0.05 in participants with CIAD8 versus No-CIAD8. Preoperative PROMs in participants with versus without probable CI on the AD8. Values are expressed as mean ± SE or n (%) where appropriate. Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; CIAD8, probable cognitive impairment as classified by the AD8; EQ−5D-5L, EuroQol 5-dimension Questionnaire; GDS, Geriatric Depression Scale; IADL, Instrumental Activities of Daily Living Scale; PROMs, patient-reported outcome measures; PSQI, Pittsburgh Sleep Quality Index; SQS, Single-item Sleep Quality Scale; VAS, visual analog scale. aIADL scores range from 0 to 8, with a lower score indicating greater dependency in functioning. bVAS pain scores range from 0 to 10, with a higher score indicating a higher pain level. c EQ−5D-5L scores range from −0.148–0.949, with a lower score indicating poorer health-related quality of life. dSQS scores range from 0 to 10, with a lower score indicating poor sleep quality. ** and * represent P ≤ 0.001 and P ≤ 0.05 in participants with CIAD8 versus No-CIAD8. Prevalence of functional disability in participants with and without probable CI on the AD8 Compared with No-CIAD8 participants, the CIAD8 group exhibited a significantly higher prevalence of functional disability and clinically significant functional disability preoperatively and at 30, 90, and 180 days postoperatively (Fig. 1A, Table S3, Fig. S2A). Specifically, the CIAD8 group reported greater difficulties across WHODAS domains of cognition, self-care, mobility, getting along with others, life activities, and participation in society at all time points, except self-care at 90 days (Table S3). Sensitivity analysis showed that the prevalence of functional disability remained significantly higher in the CIAD8 group at all perioperative time points compared with the No-CIAD8 group, after adjusting for age, sex, education level, and ASA class (Table S4, Fig. S3).Fig. 1Prevalence (A) and trajectory (B) of functional disability in participants with and without probable CI on the AD8. Error bars represent standard errors (SE). A higher score on the World Health Organization Disability Assessment Schedule (WHODAS) indicates greater functional disability (total WHODAS scores were converted to and reported as percentages by dividing the sum by 48 and multiplying by 100). Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; CIAD8, probable cognitive impairment as classified by the AD8; WHODAS, World Health Organization Disability Assessment Schedule. **represents P ≤ 0.001 in participants with CIAD8 versus No-CIAD8. †† and † represent P ≤ 0.001 and P ≤ 0.05 compared with the preoperative time point, respectively. Panel A: Compared with No-CIAD8 participants, the CIAD8 group showed a significantly higher perioperative prevalence of functional disability. While the prevalence of functional disability decreased by 16% in both groups at 180 days compared with preoperative baseline, this reduction reached statistical significance only in the No-CIAD8 group. Panel B: Compared with preoperative assessments, both the CIAD8 and No-CIAD8 groups showed significant improvement in functional disability levels at 90 and 180 days postoperatively. Prevalence (A) and trajectory (B) of functional disability in participants with and without probable CI on the AD8. Error bars represent standard errors (SE). A higher score on the World Health Organization Disability Assessment Schedule (WHODAS) indicates greater functional disability (total WHODAS scores were converted to and reported as percentages by dividing the sum by 48 and multiplying by 100). Abbreviations: AD8, Ascertain Dementia Eight-item Questionnaire; CIAD8, probable cognitive impairment as classified by the AD8; WHODAS, World Health Organization Disability Assessment Schedule. **represents P ≤ 0.001 in participants with CIAD8 versus No-CIAD8. †† and † represent P ≤ 0.001 and P ≤ 0.05 compared with the preoperative time point, respectively. Panel A: Compared with No-CIAD8 participants, the CIAD8 group showed a significantly higher perioperative prevalence of functional disability. While the prevalence of functional disability decreased by 16% in both groups at 180 days compared with preoperative baseline, this reduction reached statistical significance only in the No-CIAD8 group. Panel B: Compared with preoperative assessments, both the CIAD8 and No-CIAD8 groups showed significant improvement in functional disability levels at 90 and 180 days postoperatively. Both the CIAD8 and No-CIAD8 groups showed comparable changes in the prevalence of functional disability across all time points (P for interaction = 0.64). Compared with preoperative baseline, the No-CIAD8 group showed a significant decrease in the prevalence of functional disability at 180 days postoperatively (21% [95% CI 15, 28] vs. 37% [95% CI 31, 43], P ≤ 0.001). In contrast, the reduction in the prevalence of functional disability at 180 days did not reach statistical significance in the CIAD8 group (61% [95% CI 43, 76] vs. 77% [95% CI 64, 86], P = 0.14) (Fig. 1A, Fig. S3A). Nevertheless, 16% of CIAD8 participants became disability-free by 180 days postoperatively. Similarly, both the CIAD8 and No-CIAD8 groups demonstrated comparable changes in the prevalence of clinically significant functional disability across all time points (P for interaction = 0.21). By 180 days postoperatively, 13% of CIAD8 participants and 4% of No-CIAD8 participants no longer had clinically significant functional disability (Table S3, Fig. S2A). Seventy-two percent of participants underwent orthopedic or spinal procedures aimed at improving function. Similar findings were observed within this subgroup: the CIAD8 group demonstrated a higher prevalence of functional disability than the No-CIAD8 group at all perioperative time points (Fig. S4A). For the CIAD8 group, the prevalence of functional disability at 180 days postoperatively did not differ significantly from preoperative baseline, whereas the No-CIAD8 group showed a significant reduction. Trajectory of functional disability in participants with and without probable CI on the AD8 Compared with preoperative assessments, both the CIAD8 and No-CIAD8 groups showed significant improvement in functional disability levels at 90 and 180 days postoperatively, at comparable rates across all time points (P for interaction = 0.20) (Fig. 1B). By 180 days, 42% (95% CI 27, 60) of CIAD8 participants and 34% (95% CI 27, 42) of No-CIAD8 participants achieved a clinically important improvement in functional disability levels while 39% (95% CI 24, 57) and 57% (95% CI 49, 65) remained relatively stable, respectively (Table S3). Sensitivity analysis showed that functional disability levels at 180 days were no longer significantly different from preoperative assessments in the CIAD8 group after adjusting for age, sex, education level, and ASA class, suggesting a return to preoperative levels (Fig. S3B). Similar patterns were observed when functional disability scores were restricted to non-cognition-related WHODAS items, with both the CIAD8 and No-CIAD8 groups exhibiting comparable patterns across all time points (P for interaction = 0.70) (Table S3, Fig. S5A). The association between preoperative AD8 performance and functional disability was further assessed. Poorer performance on the AD8 was significantly associated with greater functional disability, with each unit increase in AD8 score associated with a 4% (95% CI 3, 5) increase in functional disability level across all time points (Table S5). Adverse clinical outcomes in participants with and without probable CI on the AD8 Compared with the No-CIAD8 group, the odds of postoperative delirium (OR: 7.37 [95% CI 2.25, 25.97], P = 0.001) and 180-day ED visits (OR: 5.05 [95% CI 1.34, 18.92], P = 0.014) were significantly higher in the CIAD8 group after age adjustment (Table 3, Table S6). Surgical complications were the most common reason for 180-day ED visits (n = 7). No significant differences were observed in 90-day clinical outcomes between participants with and without CIAD8 (Table S6).Table 3Adverse postoperative outcomes in participants with versus without probable CI on the AD8.Clinical outcome30 days postoperatively180 days postoperativelyNo-CIAD8(n = 255)CIAD8(n = 52)Multivariable Analysis*No-CIAD8(n = 255)CIAD8(n = 52)Multivariable Analysis*Estimate (95% CI)PEstimate (95% CI)PPostoperative deliriuma5 (2.0)7 (13.5)7.37 (2.25, 25.97)0.001----All-cause complications41 (16.1)13 (25.0)1.61 (0.76, 3.24)0.2017 (6.7)5 (9.6)1.44 (0.45, 3.86)0.50LOS, days1.0 [1.0–3.0]1.0 [1.0–4.0]−0.02 (−0.21, 0.16)0.81----Non-home discharge26 (10.2)9 (17.3)1.51 (0.61, 3.44)0.35----Mortality0 (0.0)0 (0.0)--1 (0.4)1 (1.9)--Other outcomes16 (6.3)1 (1.9)0.27 (0.01, 1.40)0.229 (3.5)5 (9.6)2.74 (0.81, 8.38)0.08ED visits15 (5.9)1 (1.9)0.29 (0.02, 1.50)0.245 (2.0)5 (9.6)5.05 (1.34, 18.92)0.01430-day readmission3 (1.2)0 (0.0)------Reoperation0 (0.0)0 (0.0)--4 (1.6)0 (0.0)--Composite outcomesb60 (23.5)19 (36.5)1.65 (0.85, 3.14)0.1318 (7.1)6 (11.5)1.63 (0.56, 4.15)0.33Values expressed as median [interquartile range (IQR)] or n (%) where appropriate. Estimates expressed as odds ratios for binary outcomes or beta-coefficients for the continuous outcome of LOS (log-transformed). Abbreviations: CIAD8, probable cognitive impairment as classified by the Ascertain Dementia Eight-item Questionnaire (AD8); ED, emergency department; LOS, length of hospital stay. aPostoperative delirium was assessed in 299 participants as 8 were converted to day surgery or discharged early. bComposite outcomes included all-cause complications, postoperative delirium, mortality, non-home discharge, ED visits, 30-day readmission, and reoperation. *Mortality, 30-day readmission, and reoperation were not adjusted due to low event counts. For 30-day outcomes of postoperative delirium, other outcomes (ED visits, readmission, and reoperation), and ED visits, multivariable analyses were adjusted for age due to event counts and concern for model overfitting. For 30-day outcomes of LOS, all-cause complications, non-home discharge, and composite adverse outcomes, multivariable analyses were adjusted for age, sex, and American Society of Anesthesiologists (ASA) class. For 180-day all-cause complications, other outcomes (ED visits, readmission, and reoperation), ED visits, and composite adverse outcomes, multivariable analyses were adjusted for age. P≤0.05 indicated statistical significance. Table 3Adverse postoperative outcomes in participants with versus without probable CI on the AD8.Clinical outcome30 days postoperatively180 days postoperativelyNo-CIAD8(n = 255)CIAD8(n = 52)Multivariable Analysis*No-CIAD8(n = 255)CIAD8(n = 52)Multivariable Analysis*Estimate (95% CI)PEstimate (95% CI)PPostoperative deliriuma5 (2.0)7 (13.5)7.37 (2.25, 25.97)0.001----All-cause complications41 (16.1)13 (25.0)1.61 (0.76, 3.24)0.2017 (6.7)5 (9.6)1.44 (0.45, 3.86)0.50LOS, days1.0 [1.0–3.0]1.0 [1.0–4.0]−0.02 (−0.21, 0.16)0.81----Non-home discharge26 (10.2)9 (17.3)1.51 (0.61, 3.44)0.35----Mortality0 (0.0)0 (0.0)--1 (0.4)1 (1.9)--Other outcomes16 (6.3)1 (1.9)0.27 (0.01, 1.40)0.229 (3.5)5 (9.6)2.74 (0.81, 8.38)0.08ED visits15 (5.9)1 (1.9)0.29 (0.02, 1.50)0.245 (2.0)5 (9.6)5.05 (1.34, 18.92)0.01430-day readmission3 (1.2)0 (0.0)------Reoperation0 (0.0)0 (0.0)--4 (1.6)0 (0.0)--Composite outcomesb60 (23.5)19 (36.5)1.65 (0.85, 3.14)0.1318 (7.1)6 (11.5)1.63 (0.56, 4.15)0.33Values expressed as median [interquartile range (IQR)] or n (%) where appropriate. Estimates expressed as odds ratios for binary outcomes or beta-coefficients for the continuous outcome of LOS (log-transformed). Abbreviations: CIAD8, probable cognitive impairment as classified by the Ascertain Dementia Eight-item Questionnaire (AD8); ED, emergency department; LOS, length of hospital stay. aPostoperative delirium was assessed in 299 participants as 8 were converted to day surgery or discharged early. bComposite outcomes included all-cause complications, postoperative delirium, mortality, non-home discharge, ED visits, 30-day readmission, and reoperation. *Mortality, 30-day readmission, and reoperation were not adjusted due to low event counts. For 30-day outcomes of postoperative delirium, other outcomes (ED visits, readmission, and reoperation), and ED visits, multivariable analyses were adjusted for age due to event counts and concern for model overfitting. For 30-day outcomes of LOS, all-cause complications, non-home discharge, and composite adverse outcomes, multivariable analyses were adjusted for age, sex, and American Society of Anesthesiologists (ASA) class. For 180-day all-cause complications, other outcomes (ED visits, readmission, and reoperation), ED visits, and composite adverse outcomes, multivariable analyses were adjusted for age. P≤0.05 indicated statistical significance. Adverse postoperative outcomes in participants with versus without probable CI on the AD8. Values expressed as median [interquartile range (IQR)] or n (%) where appropriate. Estimates expressed as odds ratios for binary outcomes or beta-coefficients for the continuous outcome of LOS (log-transformed). Abbreviations: CIAD8, probable cognitive impairment as classified by the Ascertain Dementia Eight-item Questionnaire (AD8); ED, emergency department; LOS, length of hospital stay. aPostoperative delirium was assessed in 299 participants as 8 were converted to day surgery or discharged early. bComposite outcomes included all-cause complications, postoperative delirium, mortality, non-home discharge, ED visits, 30-day readmission, and reoperation. *Mortality, 30-day readmission, and reoperation were not adjusted due to low event counts. For 30-day outcomes of postoperative delirium, other outcomes (ED visits, readmission, and reoperation), and ED visits, multivariable analyses were adjusted for age due to event counts and concern for model overfitting. For 30-day outcomes of LOS, all-cause complications, non-home discharge, and composite adverse outcomes, multivariable analyses were adjusted for age, sex, and American Society of Anesthesiologists (ASA) class. For 180-day all-cause complications, other outcomes (ED visits, readmission, and reoperation), ED visits, and composite adverse outcomes, multivariable analyses were adjusted for age. P≤0.05 indicated statistical significance. Participants with and without probable CI on the T-MoCA In contrast to the AD8, the prevalence of functional disability and clinically significant functional disability did not differ between the CIT-MoCA and No-CIT-MoCA groups, regardless of using cut-off ≤ 18 or 16 (Table 1, Table S3, Fig. S2B, Fig. S6A). The prevalence of functional disability at 180 days postoperatively was significantly lower compared with preoperative baseline in both groups (Fig. S6A, Fig. S7A). Similar to the AD8, both groups showed comparable rates of improvement in functional disability levels and non-cognition-related functional disability levels across all time points (Fig. S5B, Fig. S6B). CIT-MoCA was also associated with adverse clinical outcomes at 180 days postoperatively, including greater postoperative complications (OR: 3.89 [95% CI 1.57, 9.88], P = 0.004), ED visits (OR: 3.99 [95% CI 1.10, 16.09], P = 0.038), and composite adverse outcomes (OR: 3.07 [95% CI 1.28, 7.45], P = 0.012) (Table S7). Discussion Although the AD8 offers a simple, ultra-rapid (≤2 min) approach to screening for probable CI, its utility remains underexplored in perioperative settings. Using the AD8 as the cognitive screening tool, our exploratory study found that participants with CIAD8 experienced a significantly higher prevalence of functional disability than those without CIAD8 at all perioperative time points. Most importantly, we did not find evidence that probable CI modified the recovery trajectory or pattern of functional disability, regardless of using the AD8 or T-MoCA as the cognitive screening tool. Our novel, hypothesis-generating findings suggest that the AD8 can be administered virtually and may be useful as part of preoperative assessment to identify patients with probable CI who may warrant closer perioperative monitoring. These findings should be interpreted with caution, as the majority of participants were White, highly educated, and underwent procedures aimed at improving function. Although CI is associated with diminished functional independence and is increasingly recognized as a risk factor for adverse outcomes, the postoperative recovery trajectory of patient-reported functional disability has yet to be clearly characterized in this vulnerable population.2 Our exploratory study observed that participants with probable CI, regardless of the cognitive screening tool, exhibited a similar recovery trajectory or pattern in functional disability to those without probable CI. Similar findings were observed after restricting functional disability to non-cognition-related items, suggesting that these associations may extend beyond cognition-related difficulties. Notably, 16% of participants with CIAD8 became disability-free at 180 days after surgery, and 42% achieved a clinically important improvement in functional disability. As functional preservation is a valued outcome in older adults, this novel insight may help inform meaningful preoperative discussions between the healthcare team and high-risk patients regarding realistic recovery expectations and milestones.40 Consistent with a previous systematic review and meta-analysis, our exploratory analysis found that probable CI was associated with adverse clinical outcomes after surgery.2 Regardless of the screening tool, probable CI was significantly associated with 180-day ED visits, with this heightened risk likely attributable to surgical complications. Using the AD8, CIAD8 was associated with greater postoperative delirium (7-fold) and ED visits at 180 days (5-fold). These findings should be interpreted cautiously, as the observed associations are hypothesis-generating, especially given the low event counts, wide 95% CIs, and lack of multiplicity corrections. Future studies should evaluate whether individuals screening positive on the AD8 before surgery may benefit from additional perioperative resources, such as multicomponent programs aimed to prevent delirium, to optimize their surgical outcomes. Interestingly, we observed a difference between the AD8 and T-MoCA in identifying probable CI, with distinct associations with functional disability. Specifically, the prevalence of functional disability did not differ between the CIT-MoCA and No-CIT-MoCA groups at any perioperative time points. In contrast to the T-MoCA, the CIAD8 group demonstrated a 40% higher prevalence of preoperative functional disability than the No-CIAD8 group, which persisted postoperatively. The perioperative prevalence of functional disability remained higher in the CIAD8 group compared with the No-CIAD8 group, after adjusting for age, sex, education level, and ASA class. Beyond global functional disability, a positive AD8 screen was associated with greater perioperative difficulties across nearly all WHODAS domains: cognition, self-care, mobility, getting along with others, life activities, and participation in society. In contrast to these multidimensional functional impairments associated with the AD8, a positive T-MoCA screen was only associated with greater difficulties in life activities preoperatively and mobility at 90 days. The disparity between the AD8 and T-MoCA was likely attributable to their diagnostic performance in capturing various severities of CI, subjective nature of the measures, and scope of the assessments. Given the superior sensitivity (82% vs. 57%) and specificity (75% vs. 71%) of the AD8 in detecting dementia over MCI, a positive AD8 screen likely captures individuals with more severe probable CI and functional limitations.18 Conversely, the T-MoCA likely identified probable MCI which may not significantly interfere with functioning. By evaluating the impact of probable CI on daily activities over time, the AD8 may capture cognition through a more multidimensional scope that may better reflect patients’ overall functioning.17 Also, both the AD8 and WHODAS are self-reported measures that can be influenced by patient awareness and willingness to disclose health-related difficulties. Our novel findings support expanding preoperative evaluations to include cognitive screening, aligning with practice guidelines from the American Geriatrics Society, American College of Surgeons National Surgical Quality Improvement Programs, and American Society of Anesthesiologists Perioperative Brain Health Initiative.41, 42 Clinicians should also consider the importance of patient-reported outcomes and clinical outcomes when selecting the most appropriate cognitive screening instrument for preoperative assessment. Our findings suggest that the AD8, consisting of eight simple “Yes/No” questions, may warrant further investigation in environments with rapid turnover. Given its brevity, flexibility for in-person, online, or telephone administration, and associations with both functional disability and adverse clinical outcomes, the AD8 may represent a practical option for cognitive screening in preoperative settings, highlighting an area for further investigation. Limitations This study has several limitations. Due to restricted in-hospital visits during the pandemic, study assessments were conducted remotely. First, selection bias might have occurred as younger, more independent, and more computer literate participants might be recruited. Our study population was relatively homogeneous, with most participants being White, highly educated, and undergoing orthopedic or spinal surgery. Given that sociodemographic factors, particularly education level, can influence performance on cognitive screening tools, our findings may not be generalizable to individuals from more diverse racial or ethnic backgrounds, those with lower educational attainment, or other surgical populations.43 Second, 30%, 39%, and 40% of participants did not complete the WHODAS follow-up assessments at 30, 90, and 180 days, respectively, suggesting the potential risk of attrition bias. As participants who completed the study assessments may have experienced more favorable surgical recovery, the negative impact associated with probable CI may have been underestimated. Participants who did not complete the follow-up assessments tended to be older (74 [69 −77] years vs. 71 [68−76] years, P = 0.020), had fewer years of education (15 [12 −16] vs. 16 [15 −18], P < 0.001), and were more likely to have a positive T-MoCA screen (36% vs. 22%, P = 0.009), compared with those who completed the assessments. Nevertheless, they did not differ in their preoperative functional disability status. The statistical models also accounted for missing data under the MAR assumption. Third, we could not comprehensively assess cognition and functional disability using diagnostic evaluations. The AD8 is a self-administered cognitive screening tool, and the WHODAS is a self-reported tool that also includes a cognition-related domain, suggesting that individuals with subjective cognitive concerns on the AD8 may be more likely to report difficulties on the WHODAS. This risk of shared-method bias and construct overlap may inflate the observed association. Nonetheless, similar findings were observed when functional disability scores were restricted to non-cognition-related WHODAS items. Given the potential presence of residual confounding and the unclear temporal relationship between probable CI and functional disability, these findings should be interpreted as associations rather than as evidence of causality. Lastly, the goal of this report is to generate hypotheses regarding the perioperative risk associated with probable CI, and these findings should be considered exploratory. The relatively small number of participants with CIAD8 (n = 52) and the low event counts of several adverse clinical outcomes may have limited statistical power, potentially increasing the risk of type II error. Given the number of exploratory analyses performed, there is also an increased risk of type I error due to multiple testing, warranting further investigation in larger and more diverse surgical populations. Future studies should also examine the interaction between probable CI and frailty in functional recovery. Additionally, approaches such as latent class growth analysis may identify phenotypes of distinct recovery trajectories (e.g., rapid recovery vs persistent functional disability) and evaluate whether probable CI predicts membership in these subgroups. Investigating whether the association between probable CI and functional disability is partially mediated by depression and sleep quality—both potentially modifiable factors—using formal mediation analysis may help identify targets for preoperative intervention. Conclusions This exploratory multicenter study included a predominantly White and highly educated sample, with the majority undergoing orthopedic or spinal procedures aimed at improving function. In addition to greater perioperative functional disability, a positive preoperative screen on the AD8 was associated with higher odds of postoperative delirium and 180-day ED visits, although these findings should be interpreted cautiously given the low event counts and wide confidence intervals. While the prevalence of functional disability remained higher in participants with CIAD8 than those without, we did not find evidence that preoperative probable CI modified its recovery pattern. By 180 days, 16% of participants with CIAD8 became disability-free, and 42% achieved a clinically important improvement in functional disability. These findings suggest that a substantial proportion of participants with CIAD8 may experience meaningful functional improvement after surgery. The AD8 may be useful in informing patient-centered preoperative discussions and identifying older surgical patients with probable CI who may warrant closer perioperative monitoring. CRediT authorship contribution statement Frances Chung: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization. Ellene Yan: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. David He: Writing – review & editing, Methodology, Investigation, Conceptualization. Maria Carmela Tartaglia: Writing – review & editing, Methodology, Investigation, Conceptualization. Shabbir MH Alibhai: Writing – review & editing, Methodology, Investigation, Conceptualization. Linda Mah: Writing – review & editing, Methodology, Investigation, Conceptualization. David F. Tang-Wai: Writing – review & editing, Methodology, Investigation, Conceptualization. Jean Wong: Writing – review & editing, Methodology, Investigation, Conceptualization. Leif Erik Lovblom: Writing – review & editing, Formal analysis. Paras Kapoor: Writing – review & editing, Methodology, Investigation, Data curation, Conceptualization. Nina Butris: Writing – review & editing, Methodology, Investigation, Data curation, Conceptualization. Yasmin Alhamdah: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. All authors revised the article critically for important intellectual content, provided approval of the final version, and agreed to be accountable for all aspects of the work. Consent for publication Written informed consent was obtained from all participating subjects. Ethical statement The study was approved by the Research Ethics Board at the University Health Network and Mount Sinai Hospital (REB#: 20–6186.3 and 21–0123E). Funding This work was supported by Ontario Ministry of Health and Long-Term Care Innovation Fund, University Health Network Foundation (G−585005300860), 10.13039/100003305ResMed Foundation (G−575225462760), and Canadian Institutes of Health Research (CIHR) (funding reference number: FBD−195197) for the salary of Ellene Yan, Yasmin Alhamdah, Nina Butris, Paras Kapoor. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Declaration of competing interest Ellene Yan reports research support from the Canadian Institutes of Health Research (CIHR) Canada Graduate Scholarships Doctoral Award (FBD-195197). Jean Wong reports research support from the Ontario Ministry of Health and Long-Term Care Innovation Fund, Merck Inc. and is supported by a Merit Research Award from the Department of Anesthesiology and Pain Medicine, University of Toronto. LM reports grant funding from Brain Canada, Centre for Aging and Brain Health Innovation, and MOHLTC AFP Innovation Fund. Maria Carmela Tartaglia reports funding from NIH Clinical trials in AD: Janssen, Roche, Avanex, Merck, Green Valley, Biogen. Frances Chung reports research support from the Ontario Ministry of Health and Long-Term Care Innovation Grant, ResMed Foundation, University Health Network Foundation, and STOP-Bang Questionnaire proprietary to University Health Network. The remaining authors have no conflicts of interest to declare.