Skip to main content

Table 4 Cox proportional-hazards models investigating the association between eGFRcr-cys and incident all-cause dementia by genetic risk for dementia

From: Kidney function, albuminuria, and their modification by genetic factors and risk of incident dementia in UK Biobank

Genetic risk for dementia

Cases/Population

Age and sex adjusted

HR (95% CI)

p-value for interaction

Fully adjusteda

HR (95% CI)

p-value for interaction

Dementia non-APOE PRS

 Low

   ≥ 90

184/10,607

1 (Reference)

 

1 (Reference)

 

  60–89

491/21,127

1.11 (0.94–1.32)

 

1.07 (0.90–1.27)

 

   < 60

58/1,521

1.72 (1.28–2.31)

 

1.43 (1.06–1.92)

 

 Intermediate

   ≥ 90

762/31,517

1 (Reference)

 

1 (Reference)

 

  60–89

1,846/63,585

1.01 (0.92–1.10)

 

1.00 (0.92–1.09)

 

   < 60

259/4,663

1.80 (1.56–2.08)

 

1.58 (1.37–1.83)

 

 High

   ≥ 90

341/10,641

1 (Reference)

 

1 (Reference)

 

  60–89

877/21,087

1.07 (0.95–1.22)

 

1.08 (0.95–1.23)

 

   < 60

95/1,526

1.60 (1.27–2.01)

0.48

1.39 (1.10–1.74)

0.42

APOE allele

 ε3/ε3

   ≥ 90

595/37,092

1 (Reference)

 

1 (Reference)

 

  60–89

1,494/75,296

1.02 (0.93–1.13)

 

1.00 (0.90–1.10)

 

   < 60

211/5,657

1.81 (1.54–2.12)

 

1.50 (1.28–1.76)

 

 ε4/ε4 or ε3/ε4

   ≥ 90

874/16,962

1 (Reference)

 

1 (Reference)

 

  60–89

1,990/32,298

1.01 (0.93–1.09)

 

0.98 (0.91–1.07)

 

   < 60

215/2,227

1.53 (1.32–1.78)

 

1.27 (1.09–1.48)

 

 ε2/ε2 or ε3/ε2

   ≥ 90

83/7,721

1 (Reference)

 

1 (Reference)

 

  60–89

288/16,984

1.32 (1.03–1.68)

 

1.28 (1.00–1.64)

 

   < 60

49/1,306

2.71 (1.90–3.86)

0.04

2.25 (1.58–3.20)

0.04

  1. Abbreviations: APOE Apolipoprotein E, CI Confidence Interval, eGFRcr-cys Estimated Glomerular Filtration Rate Creatinine-Cystatin C Equation, HR Hazard Ratio, PRS Polygenic Risk Score
  2. aModels adjusted for age, sex, ethnicity, Townsend deprivation index, education, household income, country, smoking status, alcohol intake, body mass index, hypertension, and diabetes. The dementia non-APOE PRS analyses were also adjusted for APOE and the first 10 principal components of ancestry