Multiple imputation methods for longitudinal blood pressure measurements from the Framingham Heart Study
2003

Multiple Imputation Methods for Blood Pressure Data

Sample size: 2583 publication 10 minutes Evidence: moderate

Author Information

Author(s): Kang Terri, Kraft Peter, Gauderman W James, Thomas Duncan

Primary Institution: University of Southern California

Hypothesis

How do multiple imputation methods compare to case-wise deletion in genetic analyses of longitudinal blood pressure measurements?

Conclusion

Multiple imputation methods yield similar results to case-wise deletion but can provide more stable estimates when dealing with missing data.

Supporting Evidence

  • The propensity score and regression methods produced results similar to case-wise deletion.
  • Estimates of heritability were much lower for case-wise deletion and imputation methods compared to complete data.
  • The study highlighted the importance of considering missingness patterns in genetic analyses.

Takeaway

When researchers have missing data, using special methods can help them guess the missing information better, which can lead to more accurate results.

Methodology

The study compared case-wise deletion to two multiple imputation methods (propensity score and regression) using real and simulated data sets.

Potential Biases

Imputation methods that do not consider familial relationships may yield biased results.

Limitations

The study could not verify if the data were missing at random, and the imputation methods did not account for familial correlations in missingness.

Participant Demographics

The study analyzed data from two cohorts: the original cohort enrolled in 1948 and the offspring cohort enrolled in 1971.

Statistical Information

P-Value

0.34

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1471-2156-4-S1-S43

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication