Table 2.
Characteristics of joint cognitive trajectories
| Authors | Country (name of study) | Inclusion criteria | Sample size, mean age (SD), gender, ethnicity | Cognitive assessment; Joint variable | Follow-up | Trajectory analysis | Cognitive trajectory classes (% of sample) | Joint trajectory classes (% of sample) |
| Marioni et al. [44] | France (PAQUID) | Aged 65+; no missing data in cognition or covariates | 3,653 75.3 (6.8) y Female (58%) Ethnicity n.s. | MMSE Death records | 10 waves, 20 y | Joint latent class mixed models | 4 classes 1. High baseline cognition (51%) 2. Low baseline cognition (34%) 3. Slow decliners (11%) 4. Immediate decliners (4%) | 4 classes (death)† 1. Low incident rate (51%) 2. Medium (lower) incident rate (34%) 3. Medium (higher) incident rate (11%) 4. High incident rate (4%) |
| Marioni et al. [54] | France (PAQUID) | Aged 65+; no missing data for cognition or covariates | 2,854 77.0 (6.8) y Female (59%) Ethnicity n.s. | Composite score (general) Dementia (DSM-III-R) | 10 waves, 20 y | Joint latent class mixed models | 3 classes 1. Non-decliners (70%) 2. Moderate decliners (21%) 3. Fast decliners (9%) | 3 classes (dementia)‡ 1. Low incident rate (70%) 2. Medium incident rate (21%) 3. High incident rate (9%) |
| Robitaille et al. [41] | Sweden (OCTO-Twin) | Aged 80+; dizygotic or monozygotic twin pairs; no missing data in covariates | 702 82.9 (3.3) y Female (68%) Ethnicity n.s. | MMSE Martin vigorimeter | 5 waves, 8 y | Joint growth mixture models | 3 classes 1. High functioning (40%) 2. Moderate functioning (31%) 3. Low functioning (29%) | 3 classes (grip strength) 1. High functioning (40%) 2. Moderate functioning (31%) 3. Low functioning (29%) |
| Hu et al. [37] | China (CLHLS) | Aged 80–105; cognitive data available at all 7 waves | 6,842 93.4 (7.3) y Female (60%) Chinese | MMSE Death records | 7 waves, 16 y | Group-based trajectory models | 4 classes 1. Slow decline (53%) 2. Moderate decline (31%) 3. Progressive decline (13%) 4. Rapid decline (3%) | 4 classes (death)† 1. Slow increase (53%) 2. Moderate increase (31%) 3. Progressive increase (13%) 4. Rapid increase (3%) |
| Liu et al. [57] | US (PEP Study) | Aged 70+; preserved ADLs; no significant cognitive impairment; no terminal illness; cognitive data available at 2 + waves | 690 82.0 (11.9) y Female (65%) White, Others | MMSE Composite frailty score | 7 waves, 9 y | Group-based multi-trajectory models | 4 classes 1. No cognitive frailty (28%) 2. Slow cognitive decline (46%) 3. Rapid cognitive decline (20%) 4. Cognitive frailty (7%) | 4 classes (frailty) 1. No cognitive frailty (28%) 2. Progressive frailty (46%) 3. Progressive frailty (20%) 4. Cognitive frailty (7%) |
| Hochstetler et al. [69] | US and Canada (ADNI) | Aged 55–90; MMSE 24–30 (EMCI and LMCI) or 20–26 (AD); preserved ADLs; amyloid data available at 1 + wave; ADAS-Cog13 and FAQ data available at baseline and any subsequent waves (2 + waves) | 1,192 73.4 (7.4) y Female (57%) White, Others | ADAS-Cog13 FAQ | 4 waves, 2 y | Joint growth mixture models | 3 classes 1. Lowest baseline-Minimal change (69%) 2. Intermediate baseline-Slow worsening (18%) 3. Highest baseline-Steepest worsening (13%) | 3 classes (physical function) 1. Lowest baseline-Minimal change (69%) 2. Intermediate baseline-Slow worsening (18%) 3. Highest baseline-Steepest worsening (13%) |
n.s., not stated; MMSE, Mini-Mental State Examination; DSM, Diagnostic and Statistical Manual of Mental Disorders; ADAS-Cog13, Alzheimer’s Disease Assessment Scale–Cognitive Subscale 13-item version; FAQ, Functional Activities Questionnaire; EMCI, early mild cognitive impairment; LMCI, late mild cognitive impairment; AD, Alzheimer’s disease; ADL, activity of daily living. 1) The label of each trajectory class was transcribed verbatim from the articles. Where the labels were not stated in the article†, they were described based on baseline and rate of change in this table. 2) The order of each trajectory class was from the most advantaged to the most disadvantaged.