Skip to main content

Table 3 Naïve and corrected odds ratios for each fatty acid group/ratio and decline in Word Fluency Test (WFT)3 using two approaches and RCAL/SIMEX methods: Change in estimate (Δbias) and precision (Δprecision) compared to the naïve estimates; ARIC (1987–1998)1

From: Measurement error adjustment in essential fatty acid intake from a food frequency questionnaire: alternative approaches and methods

 

Odds Ratios (95% Confidence Intervals)

Fatty acid groups and ratios j 2

Naïve (Q1 = T)

External validation with biomarkers (M/N) (Approach A)

Internal repeat measurements (Q2/Q3) (Approach B)

   

RCAL

SIMEX

RCAL

SIMEX

6p

1.03

(0.93, 1.14)

0.98

(0.76, 1.25)

0.98

(0.83, 1.16)

1.04

(0.75, 1.21)

0.97

(0.80,1.18)

   Δbias4

__

__

0.05

c

0.05

c

-0.01

a

0.06

c

   Δprecision5

__

__

1.34

 

1.14

 

1.32

 

1.33

 

3p

1.01

(0.92, 1.12)

1.02

(0.77, 1.36)

1.01

(0.86, 1.17)

1.22

(0.90, 1.64)

1.15

(0.96,1.39)

   Δbias

__

__

-0.01

a

0.00

 

-0.21

a

-0.14

a

   Δprecision

__

__

1.45

 

1.12

 

1.50

 

1.19

 

6H

1.18

(1.06, 1.31)*

1.63

(1.26, 2.14)*

1.34

(1.14, 1.56)*

1.19

(0.48, 2.91)

1.20

(0.56,2.56)

   Δbias

__

__

-0.45

a

-0.16

a

-0.01

a

-0.02

a

   Δprecision

__

__

1.37

 

1.11

 

4.91

 

3.70

 

3H

0.85

(0.75, 0.96)*

0.73

(0.58, 0.90)*

0.80

(0.66, 0.96)*

0.80

(0.62, 1.04)

0.83

(0.68,1.02)

   Δbias

__

__

0.12

a

0.05

a

0.05

a

0.02

a

   Δprecision

__

__

1.21

 

1.14

 

1.31

 

1.17

 

6

1.03

(0.97, 1.14)

1.08

(0.82, 1.41)

1.04

(0.89, 1.22)

0.96

(0.76, 1.21)

0.96

(0.81,1.15)

   Δbias

__

__

-0.05

a

-0.01

a

0.07

c

0.07

c

   Δprecision

__

__

1.46

 

1.17

 

1.35

 

1.21

 

3

0.94

(0.85, 1.05)

0.71

(0.56, 0.90)*

0.80

(0.65, 0.97)*

0.85

(0.62, 1.17)

0.90

(0.73,1.11)

   Δbias

__

__

0.23

a

0.14

a

0.09

a

0.04

a

   Δprecision

__

__

1.30

 

1.21

 

1.53

 

1.23

 

3p/6p

1.01

(0.91, 1.11)

1.03

(0.81, 1.30)

1.03

(0.88, 1.21)

1.21

(0.95, 1.53)

1.14

(0.95,1.36)

   Δbias

__

__

-0.02

a

-0.02

a

-0.20

a

-0.13

a

   Δprecision

__

__

1.32

 

1.13

 

1.32

 

1.17

 

3H/6H

0.86

(0.76, 0.97)*

0.77

(0.64, 0.93)*

0.79

(0.65, 0.96)*

0.84

(0.65, 1.10)

0.87

(0.70,1.08)

   Δbias

__

__

0.09

a

0.07

a

0.02

a

-0.01

b

   Δprecision

__

__

1.14

 

1.16

 

1.33

 

1.21

 

3/6

0.95

(0.85, 1.06)

0.91

(0.76, 1.10)

0.92

(0.78, 1.08)

1.04

(0.85, 1.27)

1.04

(0.87,1.23)

   Δbias

__

__

0.04

a

0.03

a

-0.09

c

-0.09

c

   Δprecision

__

__

1.16

 

1.11

 

1.20

 

1.13

 
  1. 1 Q1: Food frequency questionnaire measurement at visit 1 of fatty acid group intake as % of energy intake or ratio of n-3 to n-6 groups. M: biomarker of fatty acid intake in cholesteryl ester fraction of plasma; N: biomarker of fatty acid intake in phospholipid fraction of plasma; Q2: Repeat of Q1 measured at visit 2 among a subset of the cohort; Q3: Repeat of Q1 measured at visit 3 among the surviving baseline cohort; WFT: Word Fluency Test; RCAL: Regression Calibration; SIMEX: Simulation Extrapolation.
  2. 2 See table 3.
  3. 3 Each exposure is entered into the logistic regression model as a z-score and the outcome is binary with 1: decline and 0: no decline based on the reliable change index (RCI) criterion. RCI < -1.645 constitutes clinically significant decline between the two ARIC visits 2 and 4, which were separated by 6 years. The odds ratio is interpreted as increase in risk of cognitive decline with each 1 SD increase in the fatty acid exposure.
  4. 4 Δbias = ORnaive - ORcorrected. a: naïve estimate is biased towards the null; b: naïve estimate is biased away from the null; c: naïve estimate is biased through the null. The null value for an odds ratio is 1.00.
  5. 5 Δprecision = CLRcorrected/CLRnaive. CLR or Confidence Limit Ratio is the ratio of the upper 95% confidence limit over the lower one.
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us