What do CD4 counts mean?
Posted by Henry Bauer on 2010/01/29
The level of CD4 cells in peripheral blood is a prime criterion for diagnosing AIDS (in the United States in particular) and for monitoring antiretroviral treatment. However, these applications of CD4 counts stem from the initial and unhappy coincidence that when “AIDS” appeared around 1980, the counting of immune-system cells was in its infancy. By now it is known that CD4 levels are extremely variable in healthy individuals, and that a variety of physiological conditions other than “HIV” may profoundly influence CD4 counts. There seems to be no fundamental evidential warrant for the manner in which HIV/AIDS diagnosis and treatment rely on CD4 counts. Juliane Sacher among others has pointed out that the levels of CD4 cells in peripheral blood are not a meaningful measure of immune-system status, since these cells move around the body according to where they seem to be needed [Alternative treatments for AIDS, 25 February 2008].
An obvious question: what is the range of CD4 counts in healthy individuals and in a variety of illnesses? (I’m grateful to Tony Lance for alerting me to some of the intriguing sources mentioned in the following).
One of the striking aspects of CD4 counts is how enormously they vary among individuals, including healthy individuals. Here, for example, are data from HIV-negative Senegalese:
C. Mair, S. E. Hawes, H. D. Agne, P. S. Sow, I. N’doye, L. E. Manhart, P. L. Fu, G. S. Gottlieb and N. B. Kiviat. Factors associated with CD4 lymphocyte counts in HIV-negative Senegalese individuals. Clinical and Experimental Immunology 151 (2007) 432-440
In any normal distribution, the standard deviation (s.d. or σ) describes the degree of scatter around the average (or mean) value. Only about 2/3 of a sample are within (±) 1 σ; in other words, about 1/6 are further from the mean on both the higher and the lower sides. In the Table above, among the men with mean CD4 count of 712, σ = 333, about 1 in every 6 men have CD4 counts below 379 or above 1045; and about 2% have counts more than 2σ above and below 712 , that is >1378) and <50. CD4 = 200 is about 1.5σ below the mean, which corresponds to about 6-7% (~1/15) of the sample. In other words, about 1 in every 15 healthy HIV-negative Senegalese men has CD4 counts below the 200 that, in HIV-positive people, is taken to be a sign of AIDS.
Of course, CD4 counts may not follow a normal distribution, especially at upper and lower levels; but since this article reports means and standard deviations without specifying a different distribution, the authors themselves are presuming it is normal. Moreover, a similarly wide range of CD4 counts and an approximation to normal distribution is shown in other data sets as well. For example, healthy North Indians were reported to have a mean CD4 count of 720 with σ = 273 and an actually observed range of 304-1864 among 200 individuals; 10% were below 400, consistent with a normal distribution which would have about 16% below 450 (Ritu Amatya, Madhu Vajpayee, Shweta Kaushik, Sunita Kanswal, R.M. Pandey, and Pradeep Seth. “Lymphocyte immunophenotype reference ranges in healthy Indian adults: implications for management of HIV/AIDS in India”. Clinical Immunology 112  290-5). Actual distributions for several African populations, however, show a skewing toward higher CD4 counts, which indeed seems highly plausible a priori — one might expect to see a definite lower bound to CD4 counts in healthy individuals (Williams et al., “HIV infection, antiretroviral therapy, and CD4+ cell count distributions in African populations”, J Inf. Dis. 194  1450-8).
Worth particular note is the comment in Amatya et al. that “These low counts could be due to physiological lymphopenia potentially caused by protein energy malnutrition, aging, antigenic polymorphism of the CD4 molecule, prolonged sun exposure, circadian rhythm, and circannual variation [9,10]”. The use of contraceptive pills by women has also been reported to influence CD4 counts (M. K. Maini, R. J. Gilson, N. Chavda, S. Gill, A. Fakoya, E. J. Ross, A. N. Phillips and I. V. Weller. “Reference ranges and sources of variability of CD4 counts in HIV-seronegative women and men”. Genitourin Med 72 [1996) 27-31]. Most of those circumstances do not represent illness. So CD4 counts can be low for a variety of fairly normal, not seriously health-threatening conditions. It follows that reliance on CD4 counts as diagnostic of “HIV disease” increases the danger that some unknown number of “HIV-positive” individuals are being told on the basis of laboratory tests — sometimes SOLELY on the basis of laboratory tests — that they are actually sick even though they feel and actually are healthy; and these people are then at risk of being consigned to toxic “treatment” for this imaginary illness. The risk is greatest if the blood tested for CD4 counts happens to have been drawn in the morning, or in the wrong season of the year, because CD4 counts vary appreciably with both those variables: T. G. Pagleironi et al., “Circannual variation in lymphocyte subsets, revisited”, Transfusion 34  512-6; F. Hulstaert et al., “Age-related changes in human blood lymphocyte subpopulations”, Clin. Immunol. Immunopathol. 70  152-8. Maini et al. (above) report a 60% variation during the day with lowest counts at 11 am. Yet another report describes a similarly large diurnal variation, from 820 at 8 am to 1320 at 10 pm (Bofill et al., “Laboratory control values for CD4 and CD8 T lymphocytes. Implications for HIV-1 diagnosis”, Clin. Exp. Immunol. 88  243-52).
Just as with the tendency to test “HIV-positive”, CD4 counts are influenced by demographic variables: “race, ethnic origin, age group, and gender” (Amatya et al.). Bofill et al. also report a steadily decreasing CD4 count with increasing age. The contrary has been reported, however, by Jiang et al. (“Normal values for CD4 and CD8 lymphocyte subsets in healthy Chinese adults from Shanghai”, Clinical and Diagnostic Laboratory Immunology, 11  811-3). The discrepancy may be owing to differing attitudes toward statistical significance: the raw numbers in Jiang et al. do show an increase with age for men and a decrease with age for women but, as with the data of Bofill et al. and all others, the standard deviations are so large, on the order of one third of the mean values, that differences and trends would have to be very considerable if they are to be statistically meaningful.
Again, Jiang et al. report no difference between Chinese men and women, whereas several sources cite women as having higher CD4 counts than men: in Britain (Maini et al.) and in more than dozen other countries in Africa, Asia, and Europe (Mair et al.). Caucasians have higher CD4 counts than Asians or Africans, according to Amatya et al. and Jiang et al., but not according to Maini et al.
All these variations under the influence of several factors would make the diagnostic application of CD4 counts problematic even if “HIV” or “AIDS” had been shown to be the salient influence on CD4 levels. However, just as with the tendency to test “HIV-positive”, CD4 counts may be “low” in a wide range of conditions; perhaps most relevant to HIV/AIDS, in tuberculosis and general trauma, as well as with primary immunodeficiency, early acute phases of such viral infections as influenza, or Dengue fever (Bofill et al.) or recent respiratory infections (Maini et al.).
Not only are CD4 counts dubious for diagnosis or prognosis; just as with the tendency to test “HIV-positive”, CD4 counts generate a number of conundrums if interpreted according to HIV/AIDS theory: the counts are often HIGHER rather than lower in conditions generally regarded as associated with poor health. For example, smokers have higher CD4 counts than non-smokers (Maini et al., Mair et al.) and prostitutes have higher counts than other women (Mair et al.). Another “striking paradox” is in “co-infection” with “HIV” and herpes:
“We observed no effect of HSV-2 status on viral load. However, we did observe that treatment naïve, recently HIV-1 infected adults co-infected with HSV-2+ at the time of HIV-1 acquisition had higher CD4+ T cell counts over time. If verified in other cohorts, this result poses a striking paradox, and its public health implications are not immediately clear” (emphases added; Barbour et al., “HIV-1/HSV-2 co-infected adults in early HIV-1 infection have elevated CD4+ T-Cell counts”, PLoS ONE 2(10)  e1080).
There seems to be no clear warrant for diagnosing AIDS by means of CD4 counts, which may be why other countries have not followed the US example of taking <200 as a criterion. Similarly, there seems to be no clear warrant for assessing the progress of antiretroviral treatment by means of CD4 counts. Two practical illustrations of that are the fact that CD4 counts do not correlate with (or, changes in are not predicted by) “viral load” (Rodriguez et al., JAMA, 296  1498-1506), and that the NIH Treatment Guidelines distinguish immunologic failure (no increase in CD4 counts) from virologic failure (no decrease in viral load) and from clinical progression (does the patient’s health improve?).
A somewhat related illustration of the failure of HIV/AIDS theory is that “AIDS” patients with Kaposi’s sarcoma may have quite high CD4 counts: see for example Maurer T, Ponte M, Leslie K. “HIV-Associated Kaposi’s Sarcoma with a High CD4 Count and a Low Viral Load”. N Engl J Med 357 (2007) 1352-3; Krown SE, Lee JY, Dittmer DP, AIDS Malignancy Consortium. “More on HIV-Associated Kaposi’s Sarcoma” N Engl J Med 358 (2008) 535-6; D.G. Power, P. J. Mulholland K. J. O’Byrne. “AIDS-related Kaposi’s Sarcoma in a Patient with a Normal CD4 Count”. Clinical Oncology 20 (2008) 97; Stebbing J, Powles T, Bower M. AIDS-associated Kaposi’s sarcoma associated with a low viral load and a high CD4 cell count. AIDS 22 (2008) 551-2. Mani, D., Neil, N., Israel, R., Aboulafia, D. M. “A retrospective analysis of AIDS-associated Kaposi’s Sarcoma in patients with undetectable HIV viral loads and CD4 counts greater than 300 cells/mm3”. J Int Assoc Physicians AIDS Care (Chic Ill) 8 (2009) 279-85.
But then it has also long been known that “AIDS” Kaposi’s sarcoma is not caused by HIV, it’s now attributed to KSHV or HHV-8, which just happened — by the sort of extraordinary coincidence or oddity that is so common in HIV/AIDS matters — just happened to appear at the same time among the same risk groups as “AIDS” and “HIV” did; and then just as mysteriously went a separate path, so that KS declined from about 40% of all “AIDS” case in 1982 to well under 10% from 1987 onwards (Table 30, p. 128 in The Origin, Persistence and Failings of HIV/AIDS Theory).
More sales in the offing for snake oil and Brooklyn Bridges.
This entry was posted on 2010/01/29 at 12:04 pm and is filed under antiretroviral drugs, HIV risk groups, HIV skepticism, M/F ratios. Tagged: A. Fakoya, A. N. Phillips, A. Timms, Ajay K. Sethi, Benigno Rodríguez, Bertran Auvert, Brian G. Williams, Brian R. Long, C. A. Lee, C. Mair, C. Sabin, CD4 counts, CD4 for “AIDS” diagnosis, CD4 for treatment prognosis, Christopher C. Whalen, Christopher Dye, D. G. Power, D. M. Aboulafia, D. Macdonald-Burns, D. Mani, D. P. Dittmer, David R. Bangsberg, Douglas F. Nixon, E. J. Ross, Eline L. Korenromp Eleanor Gouws, Elizabeth R. Sharp, Esper G. Kallas, F. Hulstaert, G. S. Gottlieb, George P. Schmid, H. D. Agne, Helena Tomiyama, Hong-Zhou Lu, I. Hannet, I. N’doye, I. V. Weller, J. Stebbing, J. Y. Lee, Jason D. Barbour, Jeffrey Martin, Juliane Sacher, K. J. O'Byrne, K. Leslie, Katia C. Bassichetto, Keith E. Garrison, L. E. Manhart, Laiyi Kang, M. A. Johnson, M. Bower, M. K. Maini, M. Ponte, Madhu Vajpayee, Mari Kitahata, Maria C. Abbate, Mariana M. Sauer, Michael M. Lederman, N. B. Kiviat, N. Chavda, N. Neil, P. B. A. Kernoef, P. J. Mulholland, P. L. Fu, P. S. Sow, P.V. Holland, Pradeep Seth, Qichao Pan, Qingneng Lin, R. Israel, R. J. Gilson, R.M. Pandey, Ritu Amatya, Ronald J. Bosch, S. E. Hawes, S. E. Krown, S. Gill, Scott Sieg, Shweta Kaushik, Solange M. Oliveira, Stephen L. Boswell, Steven G. Deek, Suhrida Yadavall, Sunita Kanswal, T. Maurer, T. Powles, T.G. Pagleironi, V. Deneys, Vinay K. Cheruv, W. Christopher Mathews, Weiming Jiang, Wilma Mackay, Xiaozhang Pan, Xinhua Weng, Yi-Wei Tang, Yile Xue. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.