Adjusting for Measurement Error in Fatty Acid Intake and Cognitive Decline
Author Information
Author(s): Beydoun May A, Kaufman Jay S, Ibrahim Joseph, Satia Jessie A, Heiss Gerardo
Primary Institution: Johns Hopkins University
Hypothesis
What is the impact of measurement error correction on the precision and bias of odds ratios in logistic models for essential fatty acid intake?
Conclusion
The study found that correcting for measurement error in dietary intake of essential fatty acids can significantly affect the precision and bias of odds ratios related to cognitive decline.
Supporting Evidence
- Attenuation factors for dietary exposure were estimated using two different approaches.
- Using biomarkers for validation showed different results compared to repeat measurements.
- Measurement error correction improved the precision of odds ratios in logistic regression models.
Takeaway
This study looked at how mistakes in reporting fatty acid intake can change the results of research on brain health, showing that fixing these mistakes can lead to better understanding.
Methodology
The study used data from the Atherosclerosis Risk in Communities (ARIC) study, applying two approaches for measurement error adjustment: external validation against biomarkers and internal repeat measurements.
Potential Biases
Potential biases include correlated errors between self-reported dietary intake and biomarker measurements.
Limitations
The study lacked a reference method that is more reliable than food frequency questionnaires and relied on short-term biomarkers for dietary intake.
Participant Demographics
Participants were 50 years or older, with a diverse racial and ethnic background, primarily from four US communities.
Statistical Information
Confidence Interval
95%
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website