Main

The National Health and Nutrition Examination Survey (NHANES), a representative sample of the United States population, randomly selects 5000 individuals each year to participate in the survey. Individuals who agree to participate are interviewed in the home and then are scheduled to be examined in the Mobile Examination Centers (MECs) a few weeks after the home interview. The informed consent document for the examination and the separate consent for the storage of biologic specimens for future use are presented to the participant after the interview in the home is completed. In 2003 we analyzed data from the 1999 and 2000 survey years to describe the sociodemographic predictors of the willingness of NHANES participants to sign the consent and allow their biologic specimens to be used for genetic research.1 We now provide an update on the NHANES experience with consent for storage of biologic specimens for future studies during the 4 years of the survey when DNA was collected and the 2 years when it was dropped from the protocol. We also examined the trends among sociodemographic factors associated with consent over time.

METHODS

Survey design

NHANES, a series of cross-sectional national surveys conducted since 1971 by the National Center for Health Statistics of the U.S. Centers for Disease Control and Prevention, obtains a nationally representative sample of the U.S. civilian noninstitutionalized population. The surveys use a complex, stratified, multistage probability sample design with unequal probabilities of selection to oversample certain subgroups of people, such as adolescents, African-Americans, and Mexican Americans.2 In 1999, NHANES became a continuous survey with 2-year data-release cycles. Persons participating in each survey year are representative of the U.S. population and are examined once in the MECs, with no longitudinal follow-up. Similar procedures were used to select the sample and conduct the interviews and examinations as in previous NHANES surveys.3 The surveys provide national statistics on the health and nutritional status of the U.S. population. One of the goals of the survey in the 1999 to 2002 survey was to establish a national probability sample of biologic specimens including genetic material for future research. During the 2003 to 2004 survey years, DNA collection was dropped from the protocol because of funding constraints.

Data collection for NHANES occurs at three levels: a brief household screening interview, an in-depth household survey interview, and a medical examination in a special MEC. Consent for the examination and a separate consent for the collection of the specimens for future studies were obtained after the individual was selected for participation in the study and the household interview was completed. Only participants aged 20 years or older were eligible for DNA collection in the 1999 to 2002 survey, although biologic samples for future research were collected on all age ranges with either parental or individual consent or assent. Although the consent included the possibility of extracting DNA from saliva, only whole blood was processed for DNA extraction. The NHANES survey and consent documents were approved by the National Center for Health Statistics Research Ethics Review Board.

This article presents the trends in the percentage of participants who signed the “Specimen Storage and Continuing Studies” consent form during the three 2-year cycles of the survey. There were minor changes to the consent document in 20001 and a major revision in 2003 and 2004 with the discontinuation of the collection of DNA for future genetic studies. The consent documents used for the storage of biologic specimens in 1999 and 2000 were previously published.1 The consent form used in the 2001 to 2002 survey that included DNA collection and future genetic studies and the 2003 to 2004 survey document without this collection are presented in Figures 1 and 2 (online only).

NHANES participants were informed that all the health data collected would be kept strictly private and that the staff were not allowed to discuss their participation in the survey under penalty of Federal law: Sections 308 (d) of the Public Health Service Act (42 USC 242m) and the Privacy Act of 1974 (5USC 552C). Participants who were interviewed and examined, and who agreed to storage of their specimens cannot be identified by any researcher using the NHANES public use microdata files. Data are recategorized, top coded, and/or suppressed before public release if there is any potential for identification of participants.

The purpose of collecting the stored specimens was to provide a resource for new scientific laboratory tests, including genetic tests, that were not available when the participant was examined. Therefore, the participant was informed that no specific studies were currently planned, and that researchers using their samples would not be contacting them for any additional information. They were also told that NCHS would not contact them with specific results but would provide information on the studies being conducted in a newsletter. Proposals for use of their specimens would be reviewed for scientific merit and by an institutional review board to ensure that the proposed research was ethical. Finally, they were told they could remove their specimens at any time and were given a toll-free number to obtain information.

The number of adults aged 20 or more years selected for the survey was 6401 in the 1999 to 2000 survey, 6941 in the 2001 to 2002 survey, and 6971 in the 2003 to 2004 survey. Race/ethnicity based on self-report was categorized as non-Hispanic white, non-Hispanic black, and Mexican American. Individuals who did not self-select into these categories were classified as “other.” Poverty-index was calculated by dividing total family income by the poverty threshold index adjusted for family size at the year of interview. All interviews were conducted in English or Spanish. If another language was spoken by the participant, a family member or local interpreter translated the interview from English. Samples sizes in the poverty index and years of education categories do not equal the total population because of item nonresponse for these variables.

With the addition of a separate consent document for specimen storage and continuing studies including DNA collection in 1999, interviewers expressed some concerns that they needed more training to adequately explain the purpose and potential use of the specimen repository. In 2000 the NHANES program held several training sessions for the interviewers and provided fact sheets and brochures to address the use of the bank and the protections that were in place to ensure that all studies would have adequate technical and ethical review. Interviewers reported that these materials were helpful with answering concerns and in obtaining respondent cooperation.

Participants were able to check a box on the consent form that allowed the storage of blood, urine, or saliva for future studies and sign the form but not agree to have blood or saliva kept for future studies using their genes.

Statistical analysis

These analyses are based on data used to monitor the survey operations. Rates are therefore not weighted to produce national estimates. Consent rates and 95% confidence intervals (CIs) and logistic regression analysis were computed using SAS version 9.1 (SAS Institute, Inc. Cary, NC). Differences between subpopulations or survey years were evaluated by examining P values calculated using a univariate t statistic. To correct for multiple comparisons within categories within a survey cycle or between the survey cycles, the Bonferroni correction was applied to this P value.4

All variables examined were entered into a logistic regression model to determine which factors independently contributed to an unwillingness to consent in the two study cycles that included the collection of DNA for genetic research. The other racial/ethnic group was deleted from the logistic regression because of the small sample size and the diverse composition of the group, and the language of interview was not modeled because it was not evenly distributed among the three race/ethnic groups. All other variables were included to control for the effect of age, education, or income on response, and pair-wise interactions in the model were explored using a P = .05 from the Wald chi-square to test for significance.

RESULTS

Table 1 presents the overall response rates to the household interview for each of the three data-release cycles (1999–2000, 2001–2002, and 2003–2004) for individuals aged 20 or more years. Response rates for the interview varied from 72% to 78% over the 6 years of the survey included in this report. In most survey years non-Hispanic whites were less likely to agree to be interviewed. There was no significant sex difference in interview consent rates, and in all survey years individuals aged 60 or more years were less likely to participate than younger age groups. Response rates between survey years were significantly higher in the 2001 to 2002 survey compared with the 1999 to 2000 survey, but were significantly lower in the 2003 to 2004 survey when compared with the 2001 to 2002 cycle among the total population and Mexican Americans.

Table 1 Overall interview response rates for NHANES 1999 to 2004 participants aged 20 or more years by demographic categories
Full size table

Table 2 presents the percentage of interviewed individuals who signed the “Specimen Storage and Continuing Studies” consent document. There were significant increases in the percentage of participants consenting between the 1999 to 2000 and 2001 to 2002 survey cycles in every demographic group observed except between those aged 50 to 59 years. This significant increase between survey cycles was even more dramatic between the 2001 to 2002 and 2003 to 2004 survey cycles when the mention of DNA collection for genetic studies was removed from the consent document. This significant increase was observed in every demographic category.

Table 2 Consent to future genetic research for NHANES 1999–2002, and nongenetic future research for NHANES 2003–2004 among interviewed participants aged 20 years or more by demographic categories
Full size table

In 1999 and 2000, non-Hispanic blacks compared with non-Hispanic whites, females, and those with a poverty index less than 1 were significantly less likely to consent to future research, whereas those aged 40 to 59 years were more likely to consent compared with those aged 60 or more years. In 2001 and 2002, females and those living at or above the poverty level were less likely to consent, whereas those with less than a high school education were more likely to consent compared with those with some college education and those aged 40 to 59 years compared with the participants aged 60 or more years. Although participants were given the option of signing the consent form but not checking the box allowing their genes to be used for future studies in the 1999 to 2002 survey, only 43 individuals in 1999 and 2000 and 38 individuals in 2001 and 2002 used this option and signed the form.

The removal of the mention of DNA collection for genetic studies from the consent document in 2003 and 2004 also removed the significant decrease in consent by non-Hispanic blacks and females. In-fact, non-Hispanic blacks and Mexican Americans were significantly more likely than non-Hispanic whites to consent to the storage of specimens.

To determine independent predictors for not consenting to future studies that included genetic research, the 1999 to 2002 data were modeled controlling for age, education, and poverty index (Table 3). Non-Hispanic black race/ethnicity was the strongest predictor for not consenting, followed by female sex. Individuals aged 40 to 59 years were significantly more likely to consent when compared with those age 60 or more years.

Table 3 Relative odds for lack of consent to genetic research from logistic regression in adult participants ages 20 years and more, NHANES 1999 to 2002
Full size table

DISCUSSION

NHANES is a national survey that collects biologic samples on a representative sample of the U.S. population. One of the objectives of the survey was to provide a nationally representative specimen bank of biologic specimens, including DNA, for future health studies. High response rates are important to the survey so response bias can be reduced or controlled and estimates generalizable to the U.S. population can be produced.

We first reported differences by demographic groups in acceptance of the “Specimen Storage and Continuing Studies” component of the survey based on response rates in 1999 and 2000.1 Participants were told that no results from future testing of their biologic samples would be returned to them and that they could remove their samples at any time. A toll-free number was provided on the consent form so they could call to remove their samples. The NHANES interviewers did not receive comments on not returning future test results, and no participant, to date, has ever called to remove his or her samples. Extensive training of our interviewers and the provision of answers to commonly asked questions (Appendix 1) significantly increased this response rate in every demographic category between the two cycles of the survey that contained the collection of DNA for future genetic studies. Training and information did not change the significantly lower response rates for non-Hispanic blacks compared with whites and females compared with males observed in the 1999 to 2000 and 2001 to 2002 surveys. Other studies have demonstrated that minority race/ethnic groups are less willing to participate in genetic research.5,6 The consistently lower consent rates for females during the years when consent was sought to store specimens for future genetic studies have not been seen in clinical genetic studies;6,7 however, a population-based study in Japan found that female gender and older age were factors for nonparticipation in a population-based cohort study involving genetic research.8 Because NHANES participants are selected to be representative of the general U.S. population with no selection if there was any disease or condition, the lack of a personal incentive of a clinically relevant reason to participate in the collection and storage of blood for continuing studies that include genetic studies may impact women more than men when faced with the decision to sign the consent document. A logistic regression analysis to determine independent predictors for not signing the consent document with the genetic research included demonstrated that being non-Hispanic black or female was the independent predictor for not consenting after controlling for age, education, and poverty index.

When the mention of genes and DNA was removed from the Continuing Studies consent document in 2003 and 2004, there was no longer a significant decrease in consent by non-Hispanic blacks and females. Consent also increased from 90.1% to 98.4%. This change suggests that the concept of genetic research was a cause for concern for some participants in the NHANES survey because removing words directly linking the collection to genetic research significantly increased response rates in every demographic category when the 2003 to 2004 survey was compared with the 2001 to 2002 or 1999 to 2000 surveys when genes were mentioned.

The debate on the appropriate consent for the collection and storage of biologic samples is ongoing and not restricted to only the NHANES project.6,9 Because the NHANES is a general health study with no overarching disease or condition driving the data collection, a general consent was requested with the only limitation that the future studies will look at health conditions. This authorization of unlimited future research has been examined in two recent studies with the conclusion that although research participants want to know and consent for the storage of their biologic specimens, the majority of participants studied were not concerned with the particular disease that would be studied.6,9 During the two 2-year survey cycles when DNA and genetic studies were included on the consent document, we offered our participants the option of storage of blood, urine, and saliva while not agreeing to the storage of blood to look at genes. Less than 1% of participants chose to check this box and still sign the consent form. We conclude that we could not properly articulate this option in the consent document; therefore, we plan on a simplified approach to consent for continuing studies in future cycles of the survey that includes the storage of biologic samples to be used for future heath studies including studies using genes.

The results of this current analysis of consent for specimen storage and continuing studies suggest that there is general acceptance across the demographic groups for participation in future research including genetic research in the general U.S. population. As our interviewers, who presented the consent document, gained experience with explaining this component, consent rates improved in every category examined. Although there was a 90% acceptance of future studies including genetic studies in 2001 and 2002, once the mention of the collection of DNA was removed in the 2003 to 2004 cycle consent form, consent was almost universal with the lowest consent rate in the 60 years or older age group.

The NHANES program is planning to reintroduce the collection of DNA in the 2007 to 2008 survey. The declining cost of DNA preparation has provided an opportunity to include this component in the survey. Because we now have experience with the use of our DNA and can provide examples of the type of genetic projects approved for use of the NHANES DNA samples, we hope we can continue to improve our consent rates and address any concerns that participants might have about the use of their biologic samples.10 Our experience between the 1999 to 2000 and 2001 to 2002 survey years demonstrates that interviewer training with specific information that addresses the participants' questions and concerns can significantly improve consent rates.