CD4 count at presentation for HIV care in the United States and Canada: Are those over 50 years more likely to have a delayed presentation?

We assessed CD4 count at initial presentation for HIV care among ≥50-year-olds from 1997-2007 in 13 US and Canadian clinical cohorts and compared to <50-year-olds. 44,491 HIV-infected individuals in the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) were included in our study. Trends in mean CD4 count (measured as cells/mm3) and 95% confidence intervals ([,]) were determined using linear regression stratified by age category and adjusted for gender, race/ethnicity, HIV transmission risk and cohort. From 1997-2007, the proportion of individuals presenting for HIV care who were ≥50-years-old increased from 17% to 27% (p-value < 0.01). The median CD4 count among ≥50 year-olds was consistently lower than younger adults. The interaction of age group and calendar year was significant (p-value <0.01) with both age groups experiencing modest annual improvements over time (< 50-year-olds: 5 [4 , 6] cells/mm3; ≥50-year-olds: 7 [5 , 9] cells/mm3), after adjusting for sex, race/ethnicity, HIV transmission risk group and cohort; however, increases in the two groups were similar after 2000. A greater proportion of older individuals had an AIDS-defining diagnosis at, or within three months prior to, first presentation for HIV care compared to younger individuals (13% vs. 10%, respectively). Due to the increasing proportion, consistently lower CD4 counts, and more advanced HIV disease in adults ≥50-year-old at first presentation for HIV care, renewed HIV testing efforts are needed.


Findings
We recently reported that the median CD4 count at first presentation for HIV care in the US and Canada increased from 256 (IQR: 96-455) to 317 (IQR: 135-517) from 1997 to 2007, yet remained below 350 cells/mm 3the current cut-off for initiating highly active antiretroviral therapy (HAART) [1,2]. Over the study period, there was an increase in the median age at first presentation for HIV care (from 40 to 43 years in 1997 to 2007, p < 0.01) [1]. According to the Centers for Disease Control and Prevention (CDC) 10% of the total incident HIV infections occurring in the US in 2006 were among adults ≥50-years-old [3]. Further, the prevalence of HIV infection in individuals ≥50 years of age is rapidly increasing [4,5], yet there is evidence that this older age group may not be as aware of HIV infection and the need for preventive measures and less likely to be tested and seek care early [6][7][8][9]. As this is the largest cohort collaboration of HIV-infected individuals in North America, we have conducted a new analysis that focuses on CD4 at first presentation for HIV care among patients ≥50-years-old.
We briefly describe study population and analytical methods; more details are provided in Althoff et al. [1].
All patients were enrollees in clinical care cohorts contributing to the North American Cohort Collaboration on Research and Design (NA-ACCORD) [10], a regional group of the International Epidemiological Databases to Evaluate AIDS (IeDEA) project. Each cohort's participation in NA-ACCORD was approved by the respective local institutional review boards. All 14 NA-ACCORD clinical cohorts agreed to participate in this study although one was excluded because their study population enrollment criteria restricted to those in later stages of HIV disease. These 13 clinical cohorts have clinical sites in 17 US states, Washington DC, and 3 Canadian provinces. Our primary focus was on HIV-infected adults who were ≥50 years of age and who first presented for clinical care between January 1997 and December 2007, as compared to individuals presenting at younger ages. First presentation for HIV clinical care was defined as the date (month and year) at which the first CD4 count was reported.
The first measured CD4 was our outcome of interest. The month and year in which the CD4 was measured were recorded. If there was more than one CD4 measurement in the first month at presentation for HIV care, we calculated the mean CD4 count for the month. Other information obtained at first presentation for care included self-reported year of birth, gender, race/ethnicity (as black, white, Latino and other/unknown) and HIV transmission risk group (male-to-male sex (MSM), injection drug use (IDU) including MSM/IDU, heterosexual contact and other/unknown).
Statistical comparisons of demographic and clinical characteristics across calendar dates were made using the Cochran-Armitage trend test for categorical variables or the Cuzick trend test for continuous variables. We determined the median absolute CD4 count and interquartile range (IQR) at first presentation for HIV clinical care annually from 1997 through 2007, by age group. Multivariate linear regression models were used to describe the annual trends in estimated mean CD4 count using a linear variable for year, stratified by age group and adjusting for cohort demographic and risk characteristics; 95% confidence intervals ([,]) were also estimated using these models. Sensitivity analyses were conducted by omitting participants from the Veterans Aging Cohort Study (VACS) and the HIV Research Network (HIVRN) as these two cohorts contribute ≈50% of the participants in the NA-ACCORD and the median age in the VACS was slightly older. Results with a twosided p-value of <0.05 were considered statistically significant. Analyses were conducted using SAS, version 9.
After excluded individuals contributing data during the first year that the cohort contributed data to the NA-ACCORD to remove individuals who may have been previously in care, a total of 67,961 adults received HIV clinical care at one of the participating NA-ACCORD sites between 1997 and 2007 and had complete date and CD4 measurement information. Of these, 21,983 (32%) had a prior history of antiretroviral therapy or HIV-1 RNA results and 1,487 (2%) had an AIDS-defining diagnosis recorded more than 3 months prior to the first recorded CD4 count. These individuals were excluded as they were likely to have been previously in care. Our study population consisted of 44,491 HIV-infected individuals.
The proportions of individuals who were < and ≥50years-old who first presented for HIV care each year are shown in Table 1; additional characteristics of the study population can be found in Althoff et al. [1]. From 1997-2007, the proportion of individuals presenting for HIV care who were aged ≥50 years increased from 17% to 27% (p-value < 0.01). The increase over time in median CD4 count at first presentation for care was similar in absolute magnitude in both age groups (67 cells/mm 3 and 63 cells/mm 3 from 1997 to 2007 among <50-yearolds and ≥50-year-olds, respectively). However, the ≥50year-olds had a median CD4 count of 266 cells/mm 3 , compared to 336 cells/mm 3 among <50-year-olds, in 2007.
The median CD4 count was consistently lower in the ≥50-year-olds compared to the <50-year-olds from 1997 to 2007 (Figure 1). The proportion of individuals at first presentation for HIV care who had a CD4 count ≥350 cells/mm 3 was lower in the ≥50-year-olds compared to the <50-year-olds; this proportion increased over time for both age groups.
In the multivariate analyses, the estimated annual change in CD4 count from 1997 to 2007 was higher among ≥50-year-olds years (7 [5 , 9]  Overall, the proportion of individuals who had an AIDS-defining diagnosis recorded at, or 3 months prior to, the first CD4 measurement was highest among those aged ≥50 years (< 50-year-olds: 10%; ≥50-year-olds: 13%; p-value < 0.01); in sensitivity analyses, these proportions increased (< 50-year-olds: 12%; ≥50-year-olds: 18%; p-value < 0.01). The proportions who had an AIDSdefining diagnosis at first presentation for care decreased from 1997 to 2007 in both age groups (Table 1). Older individuals had a greater proportion with an AIDS-defining diagnosis in all years, however this disparity decreased over time (Table 1); in sensitivity analyses the decreases were of less magnitude (≥50-year-olds: 20% in 1997 to 15% in 2007, p-value < 0.01; <50-year-olds: 13% in 1997 to 12% in 2007, pvalue < 0.01). Finally, among individuals who had an AIDS-defining diagnosis at first presentation for care, the proportion of older individuals who had ≥2 AIDSdefining diagnosis was similar to that of younger individuals (18% vs. 19%, p = 0.19).
Our study has three important findings: 1) the proportion of individuals at first presentation for care who are aged ≥50 years has increased over the past 11 years; 2) older individuals at first presentation of care consistently had a lower median CD4 count compared to younger individuals; and 3) a greater proportion of older individuals have an AIDS-defining diagnosis at, or within three months prior to, first presentation for HIV care compared to younger individuals.
The increase in the proportion of individuals who were ≥50 years at first presentation for care has implications for effective HIV management and survival for older infected individuals. Older individuals initiating HAART have a decreased immune response [11][12][13][14][15][16][17][18] and mortality increases with lower CD4 counts at HAART initiation [19]. In addition, older individuals at first presentation for care may have existing co-morbid conditions that may complicate HIV treatment decisions. From a public health perspective, a delay in presentation for treatment increases the risk for ongoing transmission [20][21][22][23]. These data suggest improved screening by health providers may help detect HIV infection earlier and at younger ages.
The estimated mean annual increase in CD4 count for individuals aged < and ≥50 years is small and likely of little clinical relevance as the within-patient variation in CD4 counts is~25%. More importantly, the annual median CD4 count is still well below the CD4 recommended for initiation of HAART [24]. The proportion of individuals presenting with a CD4 ≥350 cell/smm 3 increased in all age groups, however, the proportion was approximately 10% lower among ≥50-year-olds. This suggests the potential for greater HIV treatment initiation guideline adherence if effective testing and treatment interventions target older individuals.
Finally, our data suggest older individuals are entering into care with advanced HIV disease. The CDC recently reported an increase in the proportion of ≥50-year-olds in the US who had a first HIV diagnosis within a year before AIDS diagnosis compared to 30-< 50-year-olds [25]; the Public Health Agency of Canada has noted the increase among ≥50 year-olds [26,27]. Data from New York City showed the proportion of new HIV diagnoses that are concurrent with an AIDS diagnoses increased with older age [28].
There are limitations to our study, including our lack of data regarding time since seroconversion. We chose to stratify the data using a cut-off of 50 years. Although there were more than enough individuals for additional stratification at younger ages, additional stratification at older ages was not possible. While all age groups are experiencing modest improvements in CD4 count at presentation over time, older individuals have not "caught up." These data suggest that targeted renewed prevention and testing strategies are needed in all age groups, including those ≥50years-old. grant support from STERIS; Dr. Sterling reports receiving grant support from