Neuropsychiatric genetics: misclassification in linkage studies of phenotype-genotype research

Diagnosis of fetal alcohol spectrum disorders: a validity study of the fetal alcohol syndrome checklist

Fetal alcohol spectrum disorders (FASD) are a common cause of developmental disability, birth defects, and mortality. The performance characteristics of current diagnostic tools for FASD are not adequately reported. This study examines the performance characteristics of the Fetal Alcohol Syndrome Diagnostic Checklist (FASDC). In a population of 658 subjects from North Dakota, we used the FASDC score to examine the agreement between FASDC score, clinical diagnosis, and the Institute of Medicine criteria for FASD. All subjects were seen for evaluation in the genetic/dysmorphology clinics, which are funded by the state to provide genetic diagnostic services for residents of North Dakota. We compared the clinical diagnosis and the FASDC scores to determine the performance characteristics of the FASDC in the categorical diagnosis of fetal alcohol spectrum (FAS), other-FASD, and a group with No-FASD. Comparisons were made using univariate and logistic models of outcomes using both the presence and the absence of alcohol exposure or FASDC phenotype data. The FASDC performance characteristics for differentiation of the FAS group from non-FASD were excellent (accuracy 99%, sensitivity 99%, and specificity 99%). Logistic models for subjects with scores in the FASD range were differentiated with an accuracy of 82%, sensitivity 85%, and specificity 80% using the data on phenotype and exposure. We were able to delineate subjects with scores in the No-FASD range with an accuracy of 78%, sensitivity 64%, and specificity 81% without including the exposure and phenotype data by use of the other descriptive data (maternal characteristics, birth records, and demographic data) from the FASDC. All diagnostic tools should have performance characteristics assessed and available before adoption for use in clinical settings. The FASDC scores produce diagnostic groupings that approximate expert clinical judgment. The tool may be useful in other clinical settings for the diagnosis of FASD or as an FASD registry or research database.

Power and sample size calculations in the presence of phenotype errors for case/control genetic association studies

BACKGROUND

Phenotype error causes reduction in power to detect genetic association. We present a quantification of phenotype error, also known as diagnostic error, on power and sample size calculations for case-control genetic association studies between a marker locus and a disease phenotype. We consider the classic Pearson chi-square test for independence as our test of genetic association. To determine asymptotic power analytically, we compute the distribution's non-centrality parameter, which is a function of the case and control sample sizes, genotype frequencies, disease prevalence, and phenotype misclassification probabilities. We derive the non-centrality parameter in the presence of phenotype errors and equivalent formulas for misclassification cost (the percentage increase in minimum sample size needed to maintain constant asymptotic power at a fixed significance level for each percentage increase in a given misclassification parameter). We use a linear Taylor Series approximation for the cost of phenotype misclassification to determine lower bounds for the relative costs of misclassifying a true affected (respectively, unaffected) as a control (respectively, case). Power is verified by computer simulation.

RESULTS

Our major findings are that: (i) the median absolute difference between analytic power with our method and simulation power was 0.001 and the absolute difference was no larger than 0.011; (ii) as the disease prevalence approaches 0, the cost of misclassifying a unaffected as a case becomes infinitely large while the cost of misclassifying an affected as a control approaches 0.

CONCLUSION

Our work enables researchers to specifically quantify power loss and minimum sample size requirements in the presence of phenotype errors, thereby allowing for more realistic study design. For most diseases of current interest, verifying that cases are correctly classified is of paramount importance.