Correspondence

Disclosure of the Genetic Risk of Alzheimer's Disease

N Engl J Med 2010; 362:181-182January 14, 2010

Article

To the Editor:

Green et al. (July 16 issue)1 report that the disclosure of a positive apolipoprotein E (APOE) genotyping result to adult children of patients with Alzheimer's disease led to no short-term increases in depression or anxiety relative to a control group of subjects who underwent testing but to whom results were not disclosed. These findings are based on a flawed comparison. From a policy perspective, one would wish to compare persons who were tested and provided results with those who were not tested at all; no one advocates testing and withholding results. The study's use of an inappropriate control group created bias in favor of its null finding: the experience of being tested and coming close to learning one's susceptibility to Alzheimer's disease could “prime” subjects' awareness of their risk of the disease, artificially elevating levels of anxiety or depression after testing in the control group.

A reasonable approximation of the appropriate counterfactual condition is the full group of subjects at baseline (before testing). Table 1Table 1Differences in Measures of Anxiety and Depression after Testing with a Revised Control Group of All Subjects at Baseline. shows the effects of testing with the use of the revised control group. The mean scores and variances for all subjects at baseline were calculated from the data in Table 1 of the article by Green et al., whereas the means for other measures (adjusted for covariates) were based on the values reported in Table 3 of their article. Variances were imputed from reported standard errors and sample sizes.

We found significant increases in depression in eight of nine measures. Given self-selection and the counseling and education included in the authors' study, the true psychological consequences were probably even greater.

Sanford C. Gordon, Ph.D.
Dimitri Landa, Ph.D.
New York University, New York, NY
sanford.gordon@nyu.edu

No potential conflict of interest relevant to this letter was reported.

1 References

1. 1

   Green RC, Roberts JS, Cupples LA, et al. Disclosure of APOE genotype for risk of Alzheimer's disease. N Engl J Med 2009;361:245-254
   Full Text | Web of Science | Medline

Author/Editor Response

Gordon and Landa raise an interesting point, but we disagree that our control group was the wrong one. Our design was focused on isolating the effect of disclosure of genetic risk on persons who were motivated to learn about their own risk of Alzheimer's disease. A comparison group of persons who did not have an interest in their own risk of Alzheimer's disease would therefore be inappropriate, and a comparison group of persons who had such an interest and who received no information at all would have measured the effect of risk disclosure without reference to the genetic component. Gordon and Landa also suggest that our choice biased results in favor of the null hypothesis by priming subjects with an increased awareness of risk. This bias seems unlikely, because most of our subjects entered the study with an inflated sense of their risk of disease (i.e., their perceived risk was higher than that warranted by their status as first-degree relatives). After testing, clinically insignificant, minor elevations were observed in scales of depression symptoms, but not anxiety symptoms. In-depth interviews of subjects whose scores changed the most did not reveal any priming but instead referenced stressors that were not related to the risk of Alzheimer's disease. We agree that our results do not generalize to contexts in which APOE information is provided without the support of genetic counseling.

J. Scott Roberts, Ph.D.
University of Michigan School of Public Health, Ann Arbor, MI

L. Adrienne Cupples, Ph.D.
Boston University School of Public Health, Boston, MA

Robert C. Green, M.D., M.P.H.
Boston University School of Medicine, Boston, MA
rcgreen@bu.edu

Since publication of their article, the authors report no further potential conflict of interest.

Citing Articles (1)

Citing Articles

1. 1

   (2010) Current awareness in geriatric psychiatry. International Journal of Geriatric Psychiatry 25:7, i-viii
   CrossRef