Estimating sibling recurrence risk in population sample surveys

Area of Science:

• Genetic epidemiology
• Survey research methodology
• Public health

Background:

• Sibling recurrence risk (SRR) measures familial disease aggregation, crucial for identifying disease susceptibility genes.
• Traditional SRR estimation from family studies can be biased due to non-random sampling.
• Network sampling in household surveys provides a novel approach to unbiased SRR estimation.

Purpose of the Study:

• To introduce and evaluate network sampling methods for estimating sibling recurrence risk (SRR) and SRR ratio.
• To address the bias inherent in traditional SRR estimation methods from family-based studies.
• To demonstrate the utility of network sampling for population-based disease aggregation studies.

Main Methods:

• Two network sampling approaches for sibship ascertainment were analyzed: self-reporting and affected-individual reporting.
• Development of network estimators for SRR and SRR ratio, including standard error estimation.
• Application of methods using 1976 National Health Interview Survey data on sibling diabetes status.

Main Results:

• The SRR ratio for diabetes among living siblings was estimated at 5.79% (RSE 5.12%).
• Including deceased siblings, the SRR ratio for diabetes increased to 7.66% (RSE 3.76%).
• These results demonstrate the feasibility of network sampling for disease recurrence studies.

Conclusions:

• Network sampling estimators are effective for obtaining population-level estimates of SRR and SRR ratio.
• This methodology is applicable to various diseases, including diabetes, for assessing familial aggregation.
• Network sampling enhances the accuracy of familial disease risk assessment in epidemiological studies.