It is well established that the cardinal immunopathological mechanism that leads to profound immunodeficiency in HIV infection is the progressive loss of CD4 T lymphocytes . The HIV-1 viral load has been found to correlate with the rate of loss of CD4 T lymphocytes, representing the most commonly used prognostic markers in clinical practice [2,3]. The rate of loss of CD4 T lymphocytes in a significant proportion of HIV-infected individuals has been documented to be constant; however, considerable individual variation has been observed [4–9]. The rates of decline of CD4 T lymphocytes in HIV-infected individuals in Europe and north America have been reported in several populations [3,7,8]. However, studies that have compared the rate of disease progression and CD4 T lymphocyte decline in African and non-African populations living in the same geographical location have not been conclusive [5,6,8,10]. In a study in Nairobi involving sex workers, a rapid clinical progression from seroconversion to AIDS in a median period of 5 years was reported . A recent study in Uganda  reported a median time from seroconversion to AIDS of 9.4 years, which is similar to that reported in Europe and north America. Factors including age, mode of infection, concurrent infections, HIV-1 biological phenotypes and HIV-1 subtypes have been suggested to influence HIV-1 disease progression [3,10,11,13–18].
We have determined the rate of decline of CD4 T lymphocytes among HIV-1-infected individuals with unknown dates of seroconversion (seroprevalent), HIV-1-infected individuals with known dates of seroconversion (seroincident) and HIV-uninfected subjects who were followed up for up to 9 years in Dar es Salaam, Tanzania.
Materials and methods
The study subjects were recruited among hotel workers in Dar es Salaam, Tanzania from 1989 to 1998, in order to study the natural history of HIV-1 infection. After explaining the objectives of the study, hotel workers who consented were invited to participate in the study. The workers were interviewed and examined at recruitment and once per year. HIV testing was carried out after pre-test counselling, and post-test counselling was conducted with those who wanted to know their HIV test results. Lymphocyte subset determinations were carried out on individuals who were found to be HIV infected at recruitment (HIV-1 seroprevalent), on those who became HIV infected during the study period (HIV-1 seroincident) and on matched HIV-seronegative referents who gave verbal consent. Efforts were made to have two age and sex-matched HIV seronegative referents for every HIV-1-seropositive individual. The families, for financial reasons, regularly reported a death of a hotel worker back to the hotels. The closure of the hotels in 1998 resulted in the termination of the study. Over a span of 9 years, 1548 workers in three hotels comprising 163 HIV-1-seroprevalent (10.5%), 125 HIV-1-seroincident (8.1%) and 1260 HIV-seronegative (81.4%) individuals were recruited into this study of the progression of immunodeficiency (total study population).
The ethics committees of the National AIDS Control Programme of the Ministry of Health, Tanzania and Karolinska Institute, Stockholm, Sweden gave ethical approval for the study.
HIV antibodies in blood were demonstrated using Enzygnost anti-HIV-1+2 Plus enzyme-linked immunosorbent assay (Behring, Marburg, Germany). All reactive samples were examined using a Western blot assay (Genelab, USA), which was interpreted according to the World Health Organization criteria .
Lymphocyte subset determination
The immunophenotyping of lymphocyte subsets was performed using the SimulSET flow cytometry method (Immunocytometry System, Becton Dickinson, CA, USA) as described previously , and according to the standard guidelines for CD4 T-lymphocyte immunophenotyping in HIV-infected individuals .
Data analysis was performed using the Statistical Package for Social Scientists (SPSS 11.0; Norusis, SPSS Inc., Chicago, IL, USA). The distribution of the CD4 T-lymphocyte percentage was normal, whereas that of absolute CD4 T lymphocytes became normal after the removal of seven outliers (0.3%) out of 2539 measurements. CD4 T-lymphocyte slopes were obtained by fitting a linear regression model by the least square method for each patient. The projected rate of decline of CD4 T lymphocytes was obtained by fitting a linear regression line utilizing the median yearly loss of CD4 T cells for each category of HIV-1-infected individuals. Differences in the means and medians were calculated using the Student's t-test and Mann–Whitney test, respectively. One-way analysis of variance was used to compare several mean values and obtain within-person variance. The within-person coefficients of variation of CD4 T lymphocytes was obtained by dividing the square root of within-person variance (within-person standard deviation) by the mean multiplied by 100 on 55 HIV-seronegative individuals with seven CD4 lymphocyte determinations.
T-lymphocyte subsets were studied in 80.4% of the HIV-1-seroprevalent (131/163), 92.0% of the HIV-1-seroincident (115/125) and 34.6% of the HIV-seronegative individuals (436/1260), who were recruited as matched referents for the HIV-seropositive individuals, giving a total of 682 individuals (lymphocyte population). Among those, data from 94 out of 131 HIV-1-seroprevalent (71.7%), 77 out of 115 HIV-1-seroincident (67.0%) and 325 out of 436 HIV-seronegative (74.5%) individuals who had three or more CD4 T-lymphocyte determinations were used for determination of the CD4 T-lymphocyte slope (CD4 cell slope population). None of the HIV-1-seropositive individuals was on antiretroviral treatment. The mean follow-up of the CD4 cell slope population was 71.4, 52.9 and 86.0 months for HIV-1 seroprevalent, seroincident and seronegative individuals, respectively.
There were no significant differences in the median age at recruitment among the CD4 cell slope population compared with those who were not included in the analysis (Table 1). The CD4 cell slope population had significantly higher baseline CD4 T-cell counts in both HIV-1-seroincident and seroprevalent individuals when compared with those who were not included in the CD4 cell slope analysis. The median slope of the CD4 T-cell counts for the first two visits was significantly steeper for the CD4 cell slope population among the seroprevalent individuals but was not significantly different among the seroincident individuals. The median slope of the CD4 cell percentage for the first two determinations was significantly steeper for the CD4 cell slope population among the seroincident individuals but not among the seroprevalent individuals compared with those who were excluded from the analysis (Table 1). There was thus no indication of a selection for steeper decline in those lost to follow-up, bearing in mind the large standard deviation when slopes are calculated from only two CD4 cell observations.
Among the 325 HIV-seronegative referents who had three or more CD4 T-lymphocyte determinations performed, there was more variability of the CD4 T-lymphocyte counts compared with CD4 cell percentages with coefficients of variation of 34.8% and 16.7%, respectively. Slopes calculated on CD4 T-lymphocyte counts seemed to indicate that there was a trend towards lower measurements with a median of −6.0 cells per year over the study period, whereas this was not the case for the CD4 cell percentage, which was 0.1.
CD4 T-cell decline among HIV-1-seroprevalent and seroincident individuals
The median yearly loss of CD4 T lymphocytes among HIV-1-seroprevalent and seroincident individuals was almost similar, −21.5 cells/μl or −1.3% for the absolute count and percentage, respectively in the former, and −22.0 cells/μl or −1.5% for the latter. There was no significant difference in the median decline of CD4 T lymphocytes between men and women. The HIV-1-seroprevalent individuals with high initial CD4 T lymphocytes were found to have a significantly steeper decline of CD4 T-lymphocyte counts but not CD4 cell percentages than those with lower baseline counts (Pearson's correlation coefficient r = −0.470, P < 0.01; r = −0.166, P = 0.10, respectively). There was no significant age-related difference in the mean or median CD4 T-lymphocyte decline among HIV-1-seroincident individuals. However, there were few HIV-1-seroincident individuals above the age of 40 years (4/77).
Among individuals with three or more CD4 T-lymphocyte determinations 20 out of 94 HIV-1-seroprevalent (21.3%) and 15 out of 77 seroincident (19.5%) individuals died. The median baseline CD4 T-lymphocyte counts were not significantly different among those who died compared with those who survived (data not shown), whereas among seroincident individuals the baseline CD4 cell percentage was significantly lower for those who died compared with those who survived (16 versus 27%, P < 0.01). The median yearly loss of CD4 T lymphocytes was −52.0 cells/μl for seroprevalent individuals who died compared with −18.5 cells/μl for those who survived (P = 0.10). In seroincident individuals the corresponding values were −69.0 cells/μl and −15.0 cells/μl, respectively (P < 0.01). In the case of the CD4 cell percentage, the median yearly loss was −2.9% among seroprevalent individuals who died compared with −1.2% for seroprevalent individuals who survived (P < 0.01), while for seroincident individuals the values were −2.2% and −1.1%, respectively (P = 0.02). In HIV-1-seroincident individuals the mean duration since infection for those who died was 58 months.
Calculated rate of decline
Fig. 1 shows the projected decline of CD4 T-lymphocyte counts and percentages in HIV-1-seroincident individuals. The median time to reach 200 CD4 T lymphocytes/μl was 13.3 years compared with 3.9 years for those who progressed fast and died. The time to reach the CD4 cell percentage corresponding to 200 CD4 T lymphocytes (9.8%) in this cohort  was 11.8 years compared with 4.4 years for those who died (Fig. 1b).
We have followed a cohort of HIV-1-seroprevalent and seroincident hotel workers to determine the rate of decline of CD4 T lymphocytes. The rates of CD4 T-lymphocyte decline among HIV-seroprevalent and HIV-seroincident individuals were similar.
A comparison of the median CD4 T-lymphocyte slopes of the first two CD4 T-cell determinations between those with less than three CD4 cell determinations and those with three or more CD4 cell determinations did not show any tendency to a more rapid decline in the group with less than three determinations. It can therefore be implied that the exclusion of individuals with less than three CD4 T-cell determinations did not lead to a bias towards underestimating the CD4 T-cell decline.
We found a yearly median loss of −21.0 CD4 T lymphocytes/μl in prevalent HIV-1-infected individuals in our cohort, but as there was a decline of −6.0 cells/μl per year in the measurements in HIV-seronegative individuals over the period studied, the actual loss might thus be lower as measured by CD4 T-lymphocyte counts. In the present population, the CD4 T-lymphocyte percentage has been used to calculate CD4 T-lymphocyte counts . Using this relationship and the observed CD4 cell percentage, the yearly loss can be calculated as −26.9 CD4 lymphocytes/μl. This calculation utilizes the lower variability of CD4 T-lymphocyte decline using the CD4 cell percentage and values from all the 2476 analysed samples during the study.
To the best of our knowledge, this study is the first report on the decline of CD4 T lymphocytes in HIV-1-infected subjects in Africa that includes absolute numbers as well as CD4 cell percentages. A study among Gambian seroprevalent HIV-1-infected individuals with a median follow-up of 2.1 years showed a yearly decline of −2.1 CD4 cell percentage , which is higher than the decline of the CD4 cell percentage in our study (−1.3% in HIV-1-seroprevalent subjects). The rate of decline of CD4 lymphocytes in our population is similar to what has been reported in HIV-1-infected subjects in the United States and Europe [8,10,24,25]. In our study there was no sex difference in the rate of decline of CD4 T-lymphocyte counts. However, in the Women Interagency HIV Study and the Multicenter AIDS Cohort Study, women had a more rapid decline of CD4 T-lymphocyte counts than men .
Therefore, we did not find evidence that the immune deterioration is faster in HIV-1-infected individuals in sub-Saharan Africa than in Europe and the United States. Moreover, a projection using data from our cohort showed that the time from seroconversion to AIDS was approximately 13.3 years when using CD4 T-lymphocyte counts (200 cells/μl) and 11.8 years when using the corresponding CD4 cell percentage, which is similar to the global median time to AIDS defined by the CD4 T-lymphocyte counts [1–3]. Our results from a general urban working population in Tanzania are also in agreement with results from a recent population-based study in Uganda, which reported that the median time from HIV-1 seroconversion to AIDS was similar to that reported from western countries . In contrast, in a study in Nairobi, Kenya, restricted to female sex workers, the median time from seroconversion to clinical AIDS was less than 5 years .
Among the HIV-1-seroprevalent and seroincident individuals in the present study a significant proportion died and had experienced a significantly higher rate of loss of CD4 lymphocytes compared with those who survived. Similar patterns of decline of CD4 T lymphocytes have been reported elsewhere .
In the current study, serial CD4 T-lymphocyte counts in HIV-seronegative individuals were found to have more variability compared with the CD4 cell percentage, in agreement with previous studies [26,27]. Despite this limitation, it will be important to use both parameters, but bear in-mind the weakness of each when interpreting the data as it is unclear what values are more important in predicting the risk of complications to immunodeficiency.
In conclusion, the immunological deterioration of HIV-1-infected individuals in our Tanzanian population in which HIV-1 subtypes A, C and D are prevalent  is similar to that reported from the industrialized countries where HIV-1 subtype B is most common.
The authors would like to acknowledge all the hotel workers who agreed to participate in the study and all health workers who participated in the work.
Sponsorship: The study was supported by the Swedish International Development Agency (SIDA), Department of Research Cooperation (SAREC) as part of the TANSWED HIV programme.
1. Sotrel A, Dal Canto MC. HIV-1 and its causal relationship to immunosuppression and nervous system disease in AIDS: a review. Hum Pathol
2. Lyles RH, Munoz A, Yamashita TE, Bazmi H, Detels R, Rinaldo CR, et al
. Natural history of human immunodeficiency virus type 1 viremia after seroconversion and proximal to AIDS in a large cohort of homosexual men. Multicentre AIDS Cohort Study. J Infect Dis
3. Masel J, Arnaout RA, O'Brien TR, Goedert JJ, Lloyd AL. Fluctuations in HIV-1 viral load are correlated to CD4+ T-lymphocyte count during the natural course of infection. J Acquired Immune Defic Syndr
4. Demarest JF, Jack N, Cleghorn FR, Grenberg ML, Hoffman TL, Ottinger JS, et al
. Immunological and virological analysis of an acutely HIV type 1-infected patient with extremely rapid disease progression. AIDS Res Hum Retroviruses
5. Easterbrook PJ, Farzadegan H, Hoover DR, Palenicek J, Chmiel JS, Kaslow RA, et al
. Racial differences in rate of CD4 decline in HIV-1-infected homosexual men. AIDS
6. Del Amo J, Petruckevitch A, Phillips A, Johnson AM, Stephenson J, Desmond N, et al
. Disease progression and survival in HIV-1-infected Africans in London. AIDS
7. Cozzi Lepri A, Sabin CA, Phillips AN, Lee CA, Pezzotti P, Rezza G. The rate of CD4 decline as a determinant of progression to AIDS independent of the most recent CD4 count. The Italian Seroconversion Study. Epidemiol Infect
8. Alaeus A, Lidman K, Bjorkman A, Giesecke J, Albert J. Similar rate of disease progression among individuals infected with HIV-1 genetic subtypes A–D. AIDS
9. Sheppard HW, Lang W, Ascher MS, Vittinghoff E, Winkelstein W. The characterization of non-progressors: long-term HIV-1 infection with stable CD4+ T-cell levels. AIDS
10. Anastos K, Gange SJ, Lau B, Weiser B, Detels R, Giorgi JV, et al
. Association of race and gender with HIV-1 RNA levels and immunologic progression. J Acquired Immune Defic Syndr
11. Anzala AO, Nagelkerke NJ, Bwayo JJ, Holton D, Moses S, Ngugi EN, et al
. Rapid progression to disease in African sex workers with human immunodeficiency virus type 1 infection. J Infect Dis
12. Morgan D, Mahe C, Mayanja B, Okongo M, Lubega R, Whitworth JA. HIV-1 infection in rural Africa: is there a difference in median time to AIDS and survival compared with that in industrialized countries? AIDS
13. Phillips AN, Lee CA, Elford J, Webster A, Janossy G, Timms A, et al
. More rapid progression to AIDS in older HIV-infected people: the role of CD4+ T-cell counts. J Acquired Immune Defic Syndr
14. Kanki PJ, Hamel DJ, Sankale JL, Hsieh CC, Thior I, Barin F, et al
. Human immunodeficiency virus type 1 subtypes differ in disease progression. J Infect Dis
15. Anzala AO, Simonsen JN, Kimani J, Ball TB, Nagelkerke NJ, Rutherford J, et al
. Acute sexually transmitted infections increase human immunodeficiency virus type 1 plasma viremia, increase plasma type 2 cytokines, and decrease CD4 cell counts. J Infect Dis
16. Kaleebu P, Ross A, Morgan D, Yirrell D, Oram J, Rutebemberwa A, et al
. Relationship between HIV-1 Env subtypes A and D and disease progression in a rural Ugandan cohort. AIDS
17. Koot M, Keet I, Vos A, de Goede R, Ross M, Coutinho R, et al
. Prognostic value of HIV-1 syncytium-inducing phenotype for rate of CD4+ cell depletion and progression to AIDS. Ann Intern Med
18. Saag M, Hammer S, Lange J. Pathogenicity and diversity of HIV and implications for clinical management. J Acquired Immune Defic Syndr
1994, 7 (Suppl. 2)
19. WHO Global Program on AIDS. Proposed WHO criteria for interpretation of results from Western blot assay for HIV-1, HIV-2, and HTLV-HTLV-II. Wkly Epidemiol Rec
20. Urassa WK, Mbena EM, Swai AB, Gaines H, Mhalu FS, Biberfled G. Lymphocyte subset enumeration in HIV seronegative and HIV-1 seropositve adults in Dar es Salaam, Tanzania: determination of reference values in males and females and comparison of two flow cytometric methods. J Immunol Methods
21. Centers for Disease Control and Prevention. 1997 Revised guidelines for performing CD4+ T-cell determinations in persons infected with human immunodeficiency virus (HIV). MMWR
22. Sandstrom E, Urassa W, Bakari M, Swai A, Mhalu F, Biberfeld G, et al
. Percent CD4 T-lymphocytes can be used to estimate CD4 T lymphocyte counts in HIV-1 infected subjects in sub-Saharan Africa. Int J STD AIDS
23. Jaffar S, Wilkins A, Ngom PT, Sabally S, Corrah T, Bangali JE, et al
. Rate of decline of percentage CD4+ cells is faster in HIV-1 than in HIV-2 infection. J Acquired Immune Defic Syndr Hum Retrovirol
24. Margolick JB, Munoz A, Vlahov D, Solomon L, Astemborski J, Cohn S, et al
. Changes in T-lymphocyte subsets in intravenous drug users with HIV-1 infection. JAMA
25. Lang W, Perkins H, Anderson RE, Royce R, Jewell N, Winkelstein W Jr. Patterns of T lymphocyte changes with human immunodeficiency virus infection: from seroconversion to the development of AIDS. J Acquired Immune Defic Syndr
26. Taylor JM, Fahey JL, Detels R, Giorgi JV. CD4 percentage, CD4 number, and CD4 : CD8 ratio in HIV infection: which to choose and how to use. J Acquired Immune Defic Syndr
27. Burcham J, Marmor M, Dubin N, Tindall B, Cooper DA, Berry G, et al
. CD4% is the best predictor of development of AIDS in a cohort of HIV-infected homosexual men. AIDS
28. Lyamuya E, Olausson-Hansson E, Albert J, Mhalu F, Biberfeld G. Evaluation of a prototype Amplicor PCR assay for detection of human immunodeficiency virus type 1 DNA in blood samples from Tanzanian adults infected with HIV-1 subtypes A, C and D. J Clin Virol