Comparison of missing data approaches in linkage analysis
2003

Comparison of Missing Data Approaches in Linkage Analysis

Sample size: 4639 publication Evidence: moderate

Author Information

Author(s): Xing Chao, Schumacher Fredrick R, Conti David V, Witte John S

Primary Institution: Case Western Reserve University

Hypothesis

The impact of different methods for handling missing data on linkage analysis results is unclear.

Conclusion

Different methods for addressing missing values in linkage analyses of cohort studies can give substantially diverse results.

Supporting Evidence

  • Six different techniques were used to impute missing information for the traits BMI, CHL, and SBP.
  • The complete-subject approach generated the most significant p-values, indicating potential bias.
  • Methods I, II, and III showed agreement 72.7-83.4% of the time for the BMI trait.

Takeaway

When researchers study health data over time, missing information can change the results a lot, depending on how they handle that missing data.

Methodology

The study compared six methods for imputing missing data in linkage analyses of four traits from the Framingham Heart Study cohort.

Potential Biases

Individuals with missing values may not be randomly distributed, potentially leading to biased results.

Limitations

The complete-subject approach may introduce bias and lead to spurious results due to the elimination of individuals with missing values.

Participant Demographics

The sample included 4639 individuals, with 49% male and 51% female.

Statistical Information

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1471-2156-4-S1-S44

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication