Skip to main content

Table 3 Combined association of SCP and ICP with dementia among Chinese (N = 8112) and European (N = 44,849) middle-aged and older participants

From: Self- and interviewer-reported cognitive problems in relation to cognitive decline and dementia: results from two prospective studies

 

Non-SCP & Non-ICP

SCP & Non-ICP

Non-SCP &

ICP

SCP & ICP

China Health and Retirement Longitudinal Study (CHARLS): Probable dementia, OR (95% CI)a

 N

4495

2276

632

709

 Cases

425

405

100

190

 Age-adjusted model

1 (ref.)

2.03 (1.75, 2.36) *

1.47 (1.15, 1.87) *

2.89 (2.37, 3.53) *

 Multivariate model 1 (MV1)

1 (ref.)

1.47 (1.25, 1.72) *

1.18 (0.91, 1.52)

1.77 (1.42, 2.20) *

Survey of Health, Ageing and Retirement in Europe (SHARE): Diagnosed dementia, HR (95% CI)b

 N

31,259

11,152

1276

1162

 Cases

637

672

64

151

 Person-years

190,818

66,394

7424

6142

 Incidence rated

333.8

1012.1

862.1

2458.5

 Age-adjusted model

1 (ref.)

2.27 (2.03, 2.53) *

1.98 (1.53, 2.56) *

4.27 (3.56, 5.13) *

 Multivariate model 1 (MV1)

1 (ref.)

1.99 (1.77, 2.24) *

1.61 (1.24, 2.10) *

2.94 (2.42, 3.59) *

Pooled resultsc

 Multivariate model 1 (MV1)

1 (ref.)

1.72 (1.28, 2.31) *

1.38 (1.01, 1.88) *

2.29 (1.38, 3.77) *

  1. Abbreviations: SCP Self-reported cognitive problems, ICP Interviewer-reported cognitive problems
  2. Age-adjusted model: adjusted for age and age2
  3. Multivariate model 1 (MV1): CHARLS: further adjusted for gender, residence, marital status, education level, household income, smoking status, drinking status, sleep duration, BMI, depressive symptoms, ADL, and physical activity level; SHARE: further adjusted for gender, residence, marital status, education level, household income, smoking status, drinking status, BMI, depressive symptoms, ADL, vigorous physical activity, moderate physical activity, and country
  4. aOdds ratios (OR) were estimated using the logistic regression model
  5. bHazard ratios (HR) were estimated using the cox proportional hazards model
  6. cResults from the multivariate model 1 (MV1) were pooled using the random-effects model
  7. dIncidence per 100,000 person-years
  8. *P <0.05