IT HAS BEEN SUGGESTED THAT “Although the majority of AIDS cases in older Americans are a result of sexual exposure, not blood transfusions, sexual behavior in general, not to mention risky sexual behavior, has been virtually ignored.”1 The aim of the current study is to examine whether this alleged disregard by the public health field of older individuals can be empirically demonstrated by examining the degree to which they have been excluded from sexually transmitted disease (STD) risk-reduction clinical trials conducted over the last decade.
The inclusion of older persons in STD risk-reduction clinical trials is important for a number of reasons: (a) those over age 50 represent 11% of the total number of AIDS cases2; (b) the incidence of HIV and AIDS is growing faster among individuals 50 and older compared to those under 403,4; (c) among individuals who are at high risk for HIV infection, those who are 50 years or older are one-sixth as likely to use condoms, and one-fifth as likely to have been tested for HIV as individuals in their 20's5; (d) a survey of individuals aged 55–75 found that 47% were sexually active, 19% with multiple sexual partners6; (e) promiscuity may be increasing among older individuals due to increasing divorce rates, as well as the availability of erectile dysfunction medications3,7; and (f) among older individuals, STDs are more likely to go undetected because of the similarity of its symptoms to certain normal signs of aging and because of the lower likelihood of healthcare providers discussing STDs with this age group.1,3,8
In addition, older individuals undergo physiological changes that may make them more susceptible to contracting an STD and less likely to benefit from some treatments.4,9 These physiological changes include decreased immune functioning, as well as the thinning and drying of vaginal walls in older women, which increases the risk of tearing during sexual intercourse.2 For all these reasons listed above, older individuals are a high-risk age group for contracting STDs. It therefore appears that they could benefit from the inclusion of their age cohort in STD risk-reduction clinical trials.
A recent systematic review examined for the first time whether older individuals were excluded from clinical trials aimed at reducing the leading health-risk behaviors,10 as identified by Healthy People 2010, which included tobacco use, being overweight or obese, physical inactivity, substance abuse, and risky sexual behaviors.11 The review found that of these five types of health-risk behaviors, the clinical trials aimed at reducing sexual risk behaviors had the highest exclusion rate of older individuals. None of the clinical trials in this category included those over the age of 65.10 However, because this previous study was limited to the five most-cited medical journals,12 it was not possible to determine whether the complete exclusion of older individuals extended to a broader range of medical journals, such as specialty journals focusing on STD, health behaviors, or aging. In contrast to the previous review, the following study considered clinical trials published in any of the approximately 4220 English journals between January 1, 1994, and January 1, 2005, that are identified by MEDLINE.13
Also, the present systematic review will build on the previous systematic review by not only considering whether those over the age of 65 are excluded but also by examining whether age exclusion is observed when the threshold is lowered to over the age of 50. We selected age 50 because much of the HIV literature has designated those over the age of 50 as “older” and HIV-infected individuals over the age 50 report that they tend to be marginalized because of their age.2,14 The present systematic review will also expand the number of study characteristics explored. Thus, the objectives of our systematic review are to examine whether older individuals are excluded from STD risk clinical trials and if so, to determine whether it can be explained by characteristics of the clinical trials.
Materials and Methods
Data Sources/Study Selection
A MEDLINE literature search, covering the period from January 1, 1994, to January 1, 2005, was conducted for STD risk-reduction clinical trials that included targeting risky sexual behavior as one of its aims. For example, a study that sought to reduce the incidence of HIV only through education pertaining to the risks of needle sharing would not have been included. Relevant studies were identified through searches of the following keywords: sexually transmitted diseases, STD, HIV, AIDS, HIV/AIDS prevention, sexually transmitted infections, STI, sexual behavior, condom, prevention, chlamydia, gonorrhea, trichomonas, chancroid, genital herpes, and genital HPV. This search method included synonymous key words. We included all keywords selected for a meta-analysis of interventions aimed at preventing sexually transmitted infections.15
Our search focused on phases I, II, III, and IV of both controlled and randomized clinical trials that appeared in all English-language journals included in MEDLINE. Because our goal was to examine clinical trials having a research aim that was applicable to older participants, we did not include clinical trials aimed at preventing unwanted pregnancies or delaying sexual debut. In addition, to avoid counting a clinical trial more than once, we only included its first publication.
Of the 2844 clinical trials identified, we first eliminated 2477 because they did not have target reducing risky sexual behavior as one of its aims. Most of these articles focused on STD drug trials. Of the remaining 367 clinical trials, we eliminated 70 because their goal was to prevent pregnancy and/ or to delay sexual debut, and 67 because they were not original randomized-control trials. Lastly, we excluded 87 clinical trials that did not provide an upper-age limit or mean age with standard deviation from which upper age could be calculated. (In some cases, multiple reasons would have disqualified the trial; the first reason encountered is the one used for this tabulation.) Thus, 143 clinical trials met our inclusion criteria.
For all clinical trials, we coded whether there was an upper age limit for recruitment and, if so, whether that limit was below 50 or 65 years old.
When the clinical trials did not provide the upper age of their participants, it was calculated from the sample mean age and standard deviation. We calculated upper age for 26.6% of the clinical trials; this is similar to the percentage calculated in a prior study of age exclusion in clinical trials.16 Our method of determining age exclusion is conservative, because if the authors stated that they targeted participants between, say, the ages of 20 and 80, but did not indicate if in fact any individuals over age 50 were included, we coded this as a trial that did include those up to the age of 80.
The outcome variables abstracted from the clinical trials selected for this review were whether they included persons over the age of 50 and over the age of 65 in the study cohort and, if so, whether the studies provided information about percentage of older participants.
In addition, characteristics of the clinical trials were abstracted that included: publication year, number of participants, race and gender percentages of participants, funding source (government or other), setting (institutional or community), intrusiveness (with surgical or pharmacological interventions considered intrusive), whether participants had an STD at baseline, whether current or previous non-STD illness was an exclusion criterion, whether sex workers were included, length of clinical trial (as measured in days), country in which it was conducted (categorized as developed or developing country), whether the intervention had a significant effect, whether the outcome measure was psychological, behavioral, or biologic, and year of publication.
We conducted a series of bivariate logistic regression analyses to determine whether the likelihood of age exclusion was significantly associated with characteristics of the clinical trials. We also conducted a multivariate logistic regression with backward elimination in which all the clinical trial characteristics were entered simultaneously and age exclusion was the outcome. We used SAS (version 8.2, SAS Institute, Cary, NC) for all statistical analyses.
We also examined the validity of the assumptions that elders are not as likely to adhere to the study protocol and that they might contribute to negative results; both assumptions have been proposed as a potential explanation for the exclusion of older participants from clinical drug trials.17 We therefore coded the trials for whether the authors mentioned that older participants were either more or less likely to adhere to the study protocol and whether trials were either more or less likely to produce significant results when older individuals were included (that is, whether the interventions produced significant changes for the primary outcome in the expected direction at the 0.05 level).
Data abstraction by one reviewer was validated through a 10% audit of a random sample of clinical trials by another reviewer to confirm the accuracy of the abstracted data. The effective reliability of the two raters was 96%.18
Table 1 provides the characteristics of the participants in the 143 STD risk-reduction clinical trials. Overall, the clinical trials included 188,586 participants. Across the clinical trials that mentioned gender inclusion, women represented the majority (68.1%) of participants, and were excluded from 8.3% of the trials. Across the clinical trials that mentioned minority inclusion, minorities represented 27.7% of the participants, and were excluded from 10.1% of the clinical trials. The proportion of women and minorities was indicated for 93.7% and 80.4% of the trials, respectively.
In contrast to the low exclusion of women and minorities, 72.7% of the clinical trials excluded persons over the age of 50; the exclusion rate increased to 88.8% for those over the age of 65.
The clinical trials did not provide reasons for the exclusion of older participants. Furthermore, none of the clinical trials were limited to individuals over the age of 50, whereas 25.2% were limited to those under the age of 21. We found that 47.1% of the clinical trials that excluded individuals over the age of 50 had a stated upper age limit of 50 or younger. Also, none of the clinical trials reported the percentage of participants over the age of 65 nor did the clinical trials provide information about the gender or race distribution of the older participants.
Two variables significantly differed between those clinical trials that excluded and those that included individuals over the age of 50: government funding and type of outcome. Clinical trails that received government funding were less likely to exclude those over the age of 50, OR = 0.08 (95% CI = 0.01–0.65). In addition, clinical trials that had a biologic outcome were significantly less likely to exclude those over the age of 50, OR = 0.42 (95% CI = 0.20–0.89). In contrast, clinical trials that had a psychological [OR = 1.74 (95% CI = 0.82–3.71)] or behavioral outcome [OR = 1.34 (95% CI = 0.62–2.89)] did not differ in their tendency to exclude or include those over the age of 50. The variables of government funding and biologic outcomes were not significantly correlated, r = 0.02, P = 0.86.
The rest of the variables we considered did not significantly differ between the clinical trials that excluded and those that included individuals over the age of 50. That is, the clinical trials were equally likely to exclude those over the age of 50 regardless of: number of participants, OR = 0.91 (95% CI = 0.81–1.02); percentage of minority participants included, OR = 0.99 (95% CI = 0.99–1.01); percentage of women included, OR = 1.01 (95% CI = 0.99–1.01); setting, OR = 0.41 (95% CI = 0.08–2.01); intrusiveness, OR = 0.49 (95% CI = 0.17–1.30); whether current or previous illness was an exclusion criterion, OR = 1.32 (95% CI = 0.46–3.74); whether sex workers were included, OR = 0.49 (95% CI = 0.18–1.31); whether participants had an STD at baseline, OR = 1.00 (95% CI = 0.36–2.77); length of clinical trial, OR = 0.70 (95% CI = 0.41–1.19); or country in which it was conducted, OR = 0.84 (95% CI = 0.38–1.86).
Further, publication year did not predict the exclusion of older individuals from the clinical trials (Fig. 1). No improvement over time for inclusion of individuals over the age of 50 was found between January 1994 and January 2005, OR = 0.99 (95% CI = 0.87–1.12). Similarly, no improvement was found for inclusion of those over the age of 65 over the decade studied, OR = 1.00 (95% CI = 0.76–1.39).
When we conducted a multivariate logistic regression with all the explanatory variables entered simultaneously and backward elimination, we found the same results as when we conducted the individual bivariate logistic regression models. All of the variables dropped out of the model as being not significant, with the exception of government funding and biologic outcomes. Clinical trials that received government funding were less likely to exclude those over the age of 50, OR = 0.08 (95% CI = 0.01–0.61). In addition, clinical trials that had a biologic outcome were significantly less likely to exclude those over the age of 50, OR = 0.36 (95% CI = 0.15–0.82).
The assumption that older individuals are less likely to adhere to the study protocol was not supported by our analysis. In the eight clinical trials that mentioned an age group was more likely to adhere to the study protocol, half mentioned that older individuals were less likely to adhere and half mentioned that younger individuals were less likely to adhere. Additionally, a bivariate logistic regression showed that the inclusion of individuals over the age of 50 did not affect the likelihood of significance results.
Our findings that 72.7% of the STD risk-reduction clinical trials excluded participants over the age of 50, and 88.8% excluded those over the age of 65, are made more striking by the need for their inclusion. For example, older individuals tend to have more misconceptions about STDs than younger individuals, and receive less information and counseling about STDs from health-care providers.8,19
Exclusion from these clinical trials has the effect of suggesting to physicians and the public that older individuals are not susceptible to STDs.20 This is despite the fact that the number of AIDS cases among individuals over the age of 50 increased 500% between 1990 and 2001,21 and the proportion of older individuals who are infected by HIV is expected to increase.22 Nevertheless, our systematic review showed no improvement in the level of inclusion over the course of 10 years.
In the studies that did include older individuals, there was no indication of the proportion that was over the age of 50. In contrast 93.7% of the trials indicated the proportion of women included and 80.4% indicated the proportion of minorities included. Neither the gender nor race distribution among individuals over the age of 50 was provided by any of the clinical trials. The adoption of effective measures for this age group is hampered by not knowing whether the results of the trials applied differentially according to gender or race. To illustrate the potential usefulness of this data, almost two-thirds of the HIV cases among men over the age of 50 occur in blacks.3,23
Approximately half (52.9%) of the clinical trials did not state that they were excluding participants who were over the age of 50. Further, we found that numerous study characteristics, such as whether current or previous illness was an exclusion criterion, did not predict the likelihood of excluding individuals over the age of 50. Together, these findings suggest that an accentuated recruitment of older individuals could help to overcome the de facto ceiling. It has been shown that when older individuals are asked to participate in clinical trials, they are likely to agree.17
But recruiting an adequate number of older individuals, as well as designing studies that are intended to benefit them, may well require a corrective approach on several levels by means of: (a) federal agencies requiring that research proposals include participants over the age of 50 or else justify their exclusion, as is now required by NIH for children, women, and minority participants24; (b) directing the training of future STD risk-reduction clinical trial investigators, who could come from such disciplines as medicine, psychology, and public health, to issues related to STD prevalence among older individuals; (c) advocacy groups for older individuals generating increased awareness of their constituents' exclusion from STD risk-reduction clinical trials; and (d) journal editors and reviewers encouraging authors of STD risk-reduction clinical trial articles to show how their findings relate to older individuals, when they are included, and devoting special issues to such clinical trials that focus on older individuals.
Within STD risk-reduction clinical trials, older individuals continue to constitute an invisible group—either through their exclusion or a lack of information about their inclusion. Yet, their needs are highly visible. To the extent that the exclusion of older individuals from STD risk-reduction clinical trials reflects a pattern of age bias, a concerted effort is needed to overcome it.
1. Feldman MD. Sex AIDS and the elderly. Arch Intern Med 1994;54:19–20.
2. Goodroad BK. HIV and AIDS in people older than 50: a continuing concern. J Gerontol Nurs 2003;29:18–24.
3. Karlovsky M, Lebed B, Mydlo JH. Increasing incidence and importance of HIV/AIDS and gonorrhea among men aged ≥50 years in the US in the era of erectile dysfunction therapy. Scand J Urol Nephrol 2004;38:247–252.
4. Patel D, Gillespie, B., Foxman, B. Sexual behavior of older women: Results of a random digit-dialing survey of 2000 women in the United States. Sex Transm Dis 2003:216–220.
5. Stall R, Catania J. AIDS risk behaviors among late middle-aged and elderly Americans. The National AIDS Behavioral Surveys. Arch Intern Med 1994;154:57–63.
6. Allison-Ottey S, Weston C, Hennawi G, et al. Sexual practices of older adults in a high HIV prevalence environment. Md Med J 1999;48:287–291.
7. Lamberg L. New drug for erectile dysfunction boon for many, “viagravation” for some. JAMA 1998;280:867–869.
8. Skiest DJ, Keiser P. Human immunodeficiency virus infection in patients older than 50 years. A survey of primary care physicians' beliefs, practices, and knowledge. Arch Fam Med 1997;6:289–294.
9. Lieberman R. HIV in older Americans: an epidemiologic perspective. J Midwifery Womens Health 2000;45:176–182.
10. Levy B, Kosteas J, Slade MD, Myers LM. Exclusion of elderly persons from health-risk-behavior clinical trials. Prev Med 2004;39:625–629.
11. Services UDoHaH. Healthy People 2010: Understanding and Improving Health. Washington, DC: US Government Printing Office; 2000.
13. Fact Sheet Medline; 2005.
14. Linsk NL. HIV among older adults: age-specific issues in prevention and treatment. AIDS Read 2000;10:430–440.
15. Manhart LE, Holmes KK. Randomized controlled trials of individual-level, population-level, and multilevel interventions for preventing sexually transmitted infections: what has worked? J Infect Dis 2005;191 (Suppl 1):S7–S24.
16. Lee PY, Alexander KP, Hammill BG, Pasquali SK, Peterson ED. Representation of elderly persons and women in published randomized trials of acute coronary syndromes. JAMA 2001;286:708–713.
17. Butler RN, Nyberg JP. Clinical Trials and Older Persons: The Need for Greater Representation. New York, NY: International Longevity Center Issue Brief; 2002.
18. Rosenthal R, Rosnow RL. Essentials of Behavioral Research. Methods and Data Analysis. New York, NY: McGraw-Hill; 1991.
19. Schensul JJ, Levy JA, Disch WB. Individual, contextual, and social network factors affecting exposure to HIV/AIDS risk among older residents living in low-income senior housing complexes. J Acquir Immune Defic Syndr 2003;33 (Suppl 2):S138–S152.
20. Altschuler J, Katz AD, Tynan M. Developing and implementing an HIV/AIDS educational curriculum for older adults. Gerontologist 2004;44:121–126.
21. Mack KA, Ory MG. AIDS and older Americans at the end of the Twentieth Century. J Acquir Immune Defic Syndr 2003;33 (Suppl 2):S68–S75.
22. Szerlip MA, Desalvo KB, Szerlip HM. Predictors of HIV-infection in older adults. J Aging Health 2005;17:293–304.
23. Emlet CA. HIV/AIDS and aging: a diverse population of vulnerable older adults. J Hum Behav Soc Environ 2004;9:45–63.