Main

Alzheimer disease (AD) is a complex late-onset, degenerative disease known to have a substantial genetic component. There are three known genes associated with the early-onset autosomal dominant form of AD, but these genes account for < 2% of AD cases.1 The gene associated with the much more common late-onset AD, the ε4 allele of the apolipoprotein E (APOE) gene,2 is relatively common in the general population with approximately 25% of individuals carrying at least one ε4 allele. At the present time, however, genetic testing for late-onset AD is not widely performed or recommended because of the low predictive value of APOE genotyping and the lack of proven preventative options.3–6 As more AD susceptibility genes and effective prevention strategies are discovered, it is possible that susceptibility genotyping will be used to provide an individual with probabilistic information about future disease risk. AD susceptibility genotyping may also be used to identify at-risk individuals for clinical trials, which would also require disclosure of genotype results.7 AD is one of many complex disorders with a genetic component and given the inherent uncertainties of risk information and the rapidly advancing field of genetics, experts in the field have noted a need for multidisciplinary research focused on communication and risk perception in genetic testing: “The expected increasing uncertainty of the implications of the ‘-omics’ era will necessitate a true integration of the field of genetic counseling with health education and communication sciences.”8

Currently, most asymptomatic individuals who request a risk assessment for AD are offered a probability based on epidemiologically based risk estimates,17 yet epidemiologically derived risk estimates may have a minimal impact upon risk perception.9 Little is known about the impact of receiving genetic test results, specifically negative results, as part of a risk assessment for a common complex disorder like AD. Studies have shown that genetic information can be perceived as different from other types of disease risk information,10 a concept that has been termed “genetic exceptionalism.”11 Studies that have examined the impact of genetic test results on risk perception have come to different conclusions. One study used an analog design for a fictitious disease, with results indicating that positive genetic test results did not cause an increase in perceived risk compared to the family history risk.12 In a randomized controlled trial comparing the psychological impact of making or confirming a diagnosis of familial hypercholesterolemia using genetic testing in addition to family history and cholesterol testing, participants whose diagnosis was made using genetic testing perceived their diagnosis as more accurate.13 Although it did not examine family history information, another study focusing on biomarkers for smoking concluded that additional risk clarification by a genetic test was associated with an increase in perceived risk.14 To our knowledge, no clinical studies have explored whether individuals differentiate between a family history risk assessment and an assessment that incorporates the absence of a risk allele for a complex disorder such as AD.

The REVEAL study (Risk Evaluation and Education for Alzheimer Disease) is the first multicenter randomized clinical trial to evaluate the impact and efficacy of a risk assessment program using APOE genotyping for adult children of people diagnosed with AD. In this report, we examine the impact of an AD risk assessment protocol incorporating negative genetic test results compared to a family history risk assessment and examine the effect of the two different types of assessments on participants' perceptions of risk. The REVEAL study presented the unique opportunity to explore this subject by allowing a comparison of two subsets of participants who received identical numeric lifetime risk estimates of developing AD, but differed as to whether or not they received APOE genotype results. Based on the current literature, we expected that the individuals that received a risk assessment based on epidemiological family history data would experience little impact compared to those that received a genetic risk assessment.

MATERIALS AND METHODS

Overview

The REVEAL study is the first randomized controlled trial to evaluate the impact of a risk assessment, using APOE genotype disclosure, for AD. The study's methods have been described in detail in previous publications.7,15,16 Briefly, the study protocol was developed by a multidisciplinary team of experts in the fields of AD, neurology, genetics, genetic counseling, psychology, and bioethics. Development of the protocol was overseen and approved by a study External Advisory Board, as well as institutional review boards at each of the three study sites. All participants gave written informed consent. The various steps of the REVEAL Study are outlined in Fig. 1.

Fig 1
figure 1

Flowchart of overall REVEAL study design.

Full size image

Risk estimates

Risk estimates were formulated using two sources: (1) age-specific incidence curves for first-degree relatives of individuals affected by AD from a genetic epidemiology study of AD families17–18; and (2) odds ratio estimates reported in a meta-analysis encompassing 40 studies worldwide.19 Bayes' rule was used to stratify risk according to APOE genotype.16 The lifetime risk estimates ranged from 13% to 57% depending on an individual's gender and APOE genotype. During the Educational Session and the Disclosure Session of the protocol (Fig. 1), the genetic counselor explained that the affected parent's genotype would not be determined nor used in generating a risk assessment. Risk estimates were presented in oral, visual, and written formats and included in a take-home letter provided to participants. To facilitate communication with participants during the risk disclosure session, risk curves tailored to participant genotype and gender were generated. The genetic counselor met individually with participants in 30- to 60-minute risk disclosure sessions to communicate risk, provide support, and answer any questions that participants might have.

Participants and procedures

We analyzed the impact of risk assessment information among a subset of REVEAL participants: specifically, all women who received a lifetime risk estimate based on family history and gender (Family History Group), as well as women who possessed an APOE ε3/ε3 genotype and received a lifetime risk estimate based on genotype, gender, and family history (Genotype Group). Although men were also participants in this study, we chose to compare survey results of women in these two subgroups because both groups received identical lifetime risk estimates of 29% for developing AD. In the Genotype Group, this estimate was provided in the context of a risk assessment in which their APOE genotype was also disclosed.

Measures

At 6 weeks after risk disclosure, participants were asked a series of written survey questions designed to assess their perceptions of AD risk and the impact of risk information. All questionnaires were self-administered and completed on-site, although in rare instances, the questionnaire was mailed to the participant if he or she was unable to come to the site for the follow-up visit. The questions were as follows: (1) “Since having received a risk assessment for AD, would you say that your own chances of developing the disease have seemed” (1 = Much lower to 5 = Much higher); (2) “Since having received risk assessment for AD, how has your anxiety about developing the disease changed?” (1 = Much lower to 5 = Much higher); (3) “Please rate the overall impact that your risk information has had on you” (1 = Very positive to 5 = Very negative); (4) “I believe that I will someday develop AD” (1 = Strongly agree to 5 = Strongly disagree); and (5) “Did receiving a risk estimate help to remove uncertainty about your chance of developing AD?” (Yes, no, or undecided). Participants were also posed an open-ended question: “Please briefly describe the nature of the impact the information has had on you.”

Data analyses

Descriptive statistics were used to characterize the sample in terms of its demographics and baseline attitudes about personal risk of AD. A series of chi square analyses was used to examine study group differences on the outcomes of interest, with survey responses dichotomized to facilitate analyses. Covariates were not included in analyses because the two study groups did not differ with respect to age, race, mean years of education, number of relatives with AD or cognitive deficits, and baseline worry and belief about developing AD. The two groups differed in terms of median income but this covariate was not associated with the outcomes of interest.

RESULTS

Sample demographics

The subset of REVEAL participants with identical 29% lifetime risk estimates consisted of 36 women in the Family History Group and 30 women in the Genotype Group who were homozygous for APOE ε3. Demographic characteristics for both groups are provided in Table 1. Overall, the sample was middle aged, White, of high socioeconomic status, well educated, and had an average of approximately two family members with reported AD or related memory problems. Participants in the two groups did not differ at baseline with respect to worry or belief about developing AD.

Table 1 Sample demographics
Full size table

Data analyses

Questionnaire responses obtained at 6-week follow-ups for the two groups are shown in Table 2. Compared to their counterparts in the Family History Group, women receiving identical risk projections, along with a ε3/ε3 genotype result reported lower perceived risk of AD as a result of test information (73% vs. 25% reported lower risk, χ2 [1, N = 66] = 15.35; P < 0.0001), lower anxiety about developing AD as a result of test information (67% vs. 26% reported lower anxiety, χ2 [1, N = 64] = 10.39; P < 0.01), and a more positive overall impact of test information (80% vs. 36% reported positive impact, χ2 [1, N = 66] = 12.80; P < 0.001). A greater number of women in the Family History Group, as compared to the Genotype Group women, agreed with a statement that they believed they would someday develop AD (36% vs. 13% agreed, χ2 [1, N = 66] = 4.44; P < 0.05). In addition, women receiving a ε3/ε3 genotype were more likely to report that test information had helped remove uncertainty about whether they would develop AD (63% vs. 9%, χ2 [1, N = 60] = 15.35; P < 0.0001). Not all of the participants in both groups answered all five questions. Although participants also completed follow-up questionnaires at 6 months and 12 months after risk disclosure, only 6 week data are reported here because they reflect the most immediate impact of risk information. Analyses of 6 month and 12 month questionnaire data showed that patterns of response on study outcomes remained unchanged over time.

Table 2 Responses to survey items assessing perceived risk of AD and impact of risk information, by study group
Full size table

Qualitative data

Selected participants' responses on the open-ended question asking for a brief description of the overall impact of the risk assessment are shown in Table 3. Fifty-five of the 66 participants (83%) responded to the question at 6 weeks after disclosure. Responses were chosen for illustrative purposes in support of findings from the quantitative data.

Table 3 Selected responses to open-ended question asking about impact of risk information
Full size table

DISCUSSION

This is the first study to explore the impact on risk perception of incorporating negative genetic test results into a risk assessment for late-onset AD. Our findings suggest that inclusion of a negative genetic test result (i.e., absence of the ε4 allele) resulted in lower perceived risk of AD as compared to a family history–based risk assessment. Despite receiving the same numerical risk estimate as the Family History Group, the Genotype Group reported comparatively lower anxiety about developing AD and a more positive impact of test information. These findings are supplemented by our qualitative data. Several women in the Family History Group commented that risk information had little impact or that the information learned was what they expected. On the other hand, the Genotype Group reported optimism and relief after learning that they did not carry an ε4 allele.

There are several reasons as to why the Genotype Group may have reported a more beneficial impact than the Family History Group despite the fact that both groups received a 29% percent lifetime risk of developing AD. Individuals in the study may have been more interested in the genetic tests results than a numeric risk estimate from the outset and thus focused more intently on the APOE genotype. Several studies have suggested public interest in predictive testing for AD, and in susceptibility testing in general. For example, three recent studies surveyed interest in AD predictive testing, both in the general public and among at-risk individuals, and all found that a majority of respondents expressed interest in pursuing predictive testing.20–22 Another study showed that the lay public was generally optimistic about the potential benefits of genetic susceptibility testing.23 It is important to note, however, that interest in predictive testing often does not translate into actual test uptake, as was seen in the case of Huntington disease.24

The nature of the information presented as part of the risk assessment, as well as how the information was communicated by the genetic counselors, may have also contributed to participants' favorable perceptions of genetic information. The genetic counselors informed all individuals during the disclosure session that the numeric risk estimate was derived from population studies and also stressed that it was an approximation and not an exact number; therefore, some participants in both subgroups may have discounted the risk estimate. The risk estimate itself is a middle range number and was presented as a probability. It has been shown that individuals may not be comfortable with mathematics, statistics, and probabilities in genetic counseling sessions.25 Participants may have found it easier to process APOE genotype results as opposed to the probabilistic information because of the dichotomous nature of the genetic test results. Even though the results were presented as one of six possible genotypes, the information may have been interpreted as binary: either I have the risk gene and will develop the disease or I do not have the gene and will not develop the disease. The tendency for individuals undergoing genetic counseling to process risk information in a binary manner has been noted in prior genetic counseling research in both adult-onset and prenatal settings.26 Because the Genotype Group received genetic test results, they may have been able to process the information easier than the Family History Group who only received a numeric risk estimate of 29%. This notion is supported by preliminary analyses of information recall in the REVEAL study, which indicated that participants were more likely to recall their genotype than their numeric risk estimate.27

These findings may be consistent with concerns among the medical community that the public is influenced by popular media descriptions of genetics. In describing the term genetic exceptionalism, Thomas Murray states that genetic information is viewed as special by the public because it is treated as mysterious and powerful.11 The future and potential of genetic testing is often portrayed by the media in a sensational manner, which may lead to individuals placing more importance than appropriate on genetic information.

There also exists the concern that individuals who test negative for a susceptibility factor for a complex disorder such as AD may become complacent with respect to health behaviors despite the fact that they may still be at an increased risk for disease due to their family history.28 Studies on genetic testing and counseling for cancer that address this concern have reached contrasting conclusions, either showing that individuals that received a negative test result were less compliant to screening guidelines,29 or that receipt of a negative test result did not discourage engaging in recommended risk-reducing behaviors.30 In the current study, related but as yet unpublished analyses suggest (1) that ε4-negative participants and controls did not differ regarding the extent to which they engaged in health behaviors after risk disclosure, and (2) that the majority of ε4-negative participants correctly recognized that their risk of AD was still higher than people without family history of the disease.

There are some limitations to this study. Because the analysis called for the use of a subset of individuals from REVEAL study, our sample size was relatively small. Our participants are not representative of the general population at risk for AD, given that they are predominantly White, all female, and with a high level of education and socioeconomic status. Also, because the majority of the participants entered the study with the hope of learning their APOE genotype, the women in the Family History Group may have reacted more negatively because they were disappointed and felt that useful information was being withheld. Comments from the Family History Group in Table 3 allude to disappointment in not receiving genetic test results. The genetic counselors did educate the participants that all would receive a numeric risk estimate incorporating gender and family history. However, because risk disclosure sessions were not recorded, we could not assess potential counselor and/or process variables that may have distinguished presentation of risk information between the two groups in the study. Another limitation that may have affected participants' interpretation of the risk assessment information, specifically those in the Genotype Group, is the lack of knowledge of the affected parent's genotype. Although participants were aware that parental genotype was not used in generating a numerical risk estimate, some women in the Genotype Group may have been concerned that their affected parent developed AD due to genetic factors other that APOE ε4 and were thus less confident in the risk estimate. During the risk disclosure setting, the genetic counselors discussed what information was used to generate a risk estimate while also discussing the limitations of providing a risk assessment for a disease in which additional susceptibility genes remain to be identified.

This study is unique because we were able to explore the impact of incorporating genetic susceptibility information into a risk assessment for Alzheimer disease and focus specifically on how these test results affected AD risk perception compared to a risk estimate based on generalized epidemiologic considerations. Not only are the findings significant for AD, they are also relevant for other complex disorders as more susceptibility genes are discovered. Further studies on determinants of risk perception in common, complex disease are clearly necessary. It will also be worthwhile to examine how inclusion of genetic susceptibility results may influence motivation to adopt health behaviors that may reduce disease risk.