Health Education AIDS Liaison, Toronto

Low CD4+ T-cell counts: A Variety of causes and their implications to HIV and AIDS

By Matt Irwin, MD

printer friendly version

Low CD4+ T-cell counts (CD4 counts) are associated with a variety of conditions, including many viral infections, bacterial infections, parasitic infections, sepsis, tuberculosis, coccidioidomycosis, burns, trauma, intravenous injections of foreign proteins, malnutrition, over-exercising, pregnancy, normal daily variation, psychological stress, and social isolation. This paper presents a brief review of several studies documenting low CD4 counts in people who have these conditions. The low CD4 counts caused by some of these conditions often fall below 200 per cubic millimeter, which is the level needed to diagnose acquired immunodeficiency syndrome (AIDS) in someone who was previously positive for antibodies to the human immunodeficiency virus (HIV-positive). In addition to the diagnosis of AIDS, CD4 counts are regularly used to make treatment decisions, such as when to start antiretroviral medications and when to begin preventative antibiotics.

Because many of the conditions that cause low CD4 counts are common in people diagnosed HIV-positive, caution is advised regarding the use of CD4 counts to make treatment and diagnostic decisions. This is made more urgent since some of the conditions, like psychological stress, are greatly increased when people are told that their CD4 counts are low, which may compound the problem and cause the CD4 count to fall even furthur. Psychological stress and social isolation are also created by the diagnosis, HIV-positive, and by the diagnosis of AIDS, which may also affect the CD4 count. Finally, the widely accepted argument that HIV specifically targets CD4+ T-cells is also called into question, because it appears that low CD4 counts are a common reaction to many kinds of physical and psychological stressors. When several of these factors are combined, as is often the case in HIV and AIDS, extremely low CD4 counts may be a natural result.


Low CD4 T-cell counts are considered to be a marker of the progression of HIV infection and AIDS, and have been called the 'signature' of HIV (Balter 1997). Since HIV was first claimed to be the cause of AIDS in 1984, the CD4 count has been widely used to make treatment and diagnostic decisions , but the use of the CD4 count has been controversial, and recommendations regarding how to use them have changed many times over the years.

In addition to low CD4 counts, the CD4/CD8 ratio is also considered a marker of disease progression in HIV and AIDS, and is often found to be inverted. An 'inverted' ratio simply means that there are less CD4 cells than CD8 cells, resulting in a ratio of less than 1. Another finding that is common in people diagnosed HIV-positive is reduced lymphocyte activity and function, as measured by their responsiveness to foreign antigens. This can result in a state of "anergy", where people's skin fails to respond when antigens are injected under it. There are two major arms of the immune system, one which works through antibodies produced by B-cells and plasma cells, and the other that works through cells including CD4+ T-cells. The first is called antibody-mediated or humoral immunity, and the second is called cell-mediated immunity. It is the cell-mediated arm of the immune system that is found to be profoundly suppressed in people diagnosed with AIDS. The antibody-mediated arm of the immune system, however, is hyperstimulated, with "increasing levels of humoral antibodies and plasma cells" (Fox 1996). The only measurement commonly used in clinical practice, however, is the CD4 count, as the following treatment and diagnostic recommendations demonstrate (Cecil Textbook of Medicine, Goldman 2000):

  • When the CD4 count in someone diagnosed HIV-positive is found to be below 200, AIDS is diagnosed. This method currently accounts for over half of all AIDS diagnoses, and so is highly significant (CDC 1999).
  • There are two approaches regarding when to start antiretroviral therapy. The more aggressive approach recommends starting when the CD4 count falls below 500, and the second approach is to wait until it is below 350 unless the viral load is also above 20,000 copies per ml. The more aggressive approach is more commonly used in the United States, and many clinicians will even start antiretroviral medications immediately in all patients, regardless of the patient's CD4 counts.
  • To prevent pneumocystis carinii pneumonia (PCP), antibiotics should be started if the CD4 count is found to be below 200. The most commonly used combination is sulfamethoxazole/trimethoprim (SMX/TMP), commonly referred to by its brand name, Bactrim.
  • To prevent fungal infections, the antifungal medication fluconazole should be started if the CD4 count is below 200.
  • To prevent mycobacterium avium complex (MAC) infection, the antibiotics clarithromycin, azithromycin, or rifabutin should be started if the CD4 count is below 100.
  • To prevent cytomegalovirus (CMV) infection, oral gancyclovir can be started, although no CD4 level or other guideline is given.
  • "Over the years, the recommendations of when to begin therapy for HIV have fluctuated back and forth, and a prior trend to treat most patients with fewer than 500 cells/mm3 with zidovudine (AZT) was modified by the results of a large randomized study (the Concorde Trial) showing that early AZT therapy did not yield improvement in survival." (Goldman 2000, page 1939)

Recent media reports have stated that new recommendations from the National Institute of Health will be presented in February, 2001, calling for a halt to the widespread practice of starting antiretroviral medications immediately, regardless of the patient's health status or CD4 count (Garrett 2001). This marks a retreat from the "hit hard, hit early" approach advocated by David Ho and others since 1996 when protease inhibitor combination therapy was begun. The change is being recommended, according to the reports, because the toxicities of the drugs and the difficulty in staying on them for long periods of time were underestimated.

While most people know about the reports of lowered CD4 levels in people diagnosed HIV-positive, which continue to receive widespread press coverage, other reports concerning lowered CD4 counts in people who are HIV-negative have been widely ignored. These reports show that CD4 counts commonly fall very low, especially if a person suffers from certain conditions. These conditions include a variety of viral illnesses, bacterial infections, parasitic infections, sepsis, septic shock, multiple organ system failure, tuberculosis, coccidioidomycosis, burns, trauma, transfusions, malnutrition, over-exercising, pregnancy, normal daily variation, psychological stress, and social isolation. In addition to lowered CD4 counts, these conditions result in other immunosuppressive changes that are also identical to those seen in people diagnosed HIV-positive, including reduced CD4/CD8 ratios, reduced lymphocyte function, anergy, atrophy of lymphoid organs, and general suppression of cell-mediated immunity. Several studies that examine these effects will be reviewed here, and studies will be emphasized if they reveal either lowered absolute CD4 counts, lowered CD4 percentages, or a reduced CD4/CD8 ratio.

1) Low CD4 counts in the intensive care unit

In 1995, Feeney et al. looked at CD4 counts in 102 consecutive intensive care unit (ICU) patients who were admitted for a variety of reasons, all of whom were HIV negative. The patients suffered from 34 different illnesses, with the most common being myocardial infarction (heart attack), severe bleeding, renal failure, trauma, and chronic pulmonary disease. 30% of these patients had CD4 counts less than 300, and 41% had CD4 counts less than 400. The authors do not discuss how many had counts below 200, the level resulting in a diagnosis of AIDS, or exactly how many had counts below 500, the level at which antiretroviral medications would be started in someone who has been diagnosed HIV-positive. They also did not find that low CD4 counts were linked with a poor prognosis. Here are the author's comments on their findings.

Our results demonstrate that acute illness alone, in the absence of HIV infection, can be associated with profoundly depressed lymphocyte concentrations. Although we hypothesized that this depression would be directly related to the severity of illness, this was not seen in our results. The T-cell depression we observed was unpredictable and did not correlate with severity of illness, predicted mortality rate, or survival rate. This study was consistent with prior studies that have shown similar decreases in T-cell counts in specific subsets of acutely ill patients. These subsets included patients with bacterial infections, sepsis, septic shock, multiple organ system failure, tuberculosis, coccidioidomycosis, viral infections, burns, and trauma patients. Most of these studies reported decreases in lymphocyte populations, some of which were severe and included CD4/CD8 ratio inversions...

In the largest study to date of hospitalized patients, Williams et al (1983) evaluated T-cell subsets in 146 febrile patients with serious acute infections... with 19 of 45 patients having a CD4 count of less than 300 per microliter.

We also found that CD4 counts were linearly related to total lymphocyte concentrations, as Blatt et al. (1991) reported in HIV-positive patients. (Feeney et al. 1995, pages 1682-1683)

These researchers did not find that low CD4 cell counts were good measures of prognosis, although some other reports differ in this regard.

2) Low CD4 counts in Various Human Infections

2a) Pneumonia, pyelonephritis, abscesses, infected wounds, cellulitis, and sepsis

In 1983, about one year before HIV was first mentioned as a possible cause of AIDS, Williams et al. published a study showing severely reduced CD4 counts in 146 consecutive people with serious acute infections who were admitted to their hospital in New Mexico. This article was referred to in the article by Feeney et al. that was reviewed above. The infections included pneumonia, acute pyelonephritis, abscesses, infected wounds, cellulitis, deep tissue infections, and sepsis.

The authors only provide average CD4 counts for the majority of patients, except for a graph on page 811 that plots the CD4 counts for all 45 pneumonia patients. This reveals that 31 of 45 (69%) had CD4 counts less than 500 cells/mm3, 19 of 45 (42%) had counts below 300, 13 of 45 (29%) had counts below 200, 6 of 45 (13%) had 100 or less, and 2 of 45 (4%) had values less than 50. The average CD4 count for all the people with pneumonia was 574. They also provide tables with clinical information and CD4 counts for 9 patients with soft tissue infections (STI) and 12 patients with sepsis/deep infections, all of whom had multiple T-cell abnormalities. Brief descriptions of all the cases from these tables who had CD4 counts less than 200 follow:

  • a 25 year-old female with "disseminated varicella", CD4 count of 58, "rapid septic course, death".
  • a 41 year-old male with Group A hemolytic streptococcal sepsis, CD4 count of 150, "rapid progression, ... death".
  • a 42 year-old male with E-coli sepsis, CD4 count of 156, "Multiple previous episodes of E-coli bacteremia".
  • a 38 year-old female with a submandibular abscess, CD4 count of 183, "Gram-positive organism, short 5-day hospital stay".
  • a 46 year-old man with peritonitis, CD4 count of 205, who was on long-term peritoneal dialysis.
  • a 58 year-old female with infected decubitous ulcers, CD4 count of 189, "prolonged 4-month hospital stay, ... E-coli and pseudomonas organisms".
  • 75 year-old female with a common duct stone and ascending cholangitis, CD4 count of 139, "multiple positive blood culture results".

These case examples are notable in that some of them sound very similar to people who die of AIDS, such as the 25 year old female who died of disseminated varicella with a CD4 count of only 58 cells/mm3. It is also remarkable that 30% of people with pneumonia, which is a very common illness in people diagnosed HIV-positive, had CD4 counts below 200.

2b) Low CD4 counts in malaria

Malaria is caused by parasites from the plasmodium species, and is extremely common in Africa and in many tropical areas. In 1999 a letter was published documenting severely lowered CD4 counts in African patients with malaria (Chirenda 1999). The author examined the CD4 count in 78 patients with malaria who were HIV-positive, and 19 who were HIV-negative. He was surprised to find that more HIV-negative malaria cases had severely lowered CD4 counts than did the HIV-positive cases, on average, with 8 of 19 (42%) HIV-negative cases being below 200, while only 31 of 78 (40%) HIV-positive cases had CD4 counts below 200. Seven HIV-negative malaria cases had CD4 counts below 100. This data comes from Table 1. In addition, 6 HIV-positive patients had normal CD4 counts, and the author states, "One may want to hypothesise that malaria reduces the CD4 count more than HIV infection". The author did not do statistical analyses to test for statistical significance, nor does he mention the general health or nutrition status of the patients, which may have contributed to their severely lowered CD4 counts, as will be reviewed later in this paper.

2c) Low CD4 counts in mononucleosis

Mononucleosis, commonly called 'mono', is a common viral illness, especially in young people of college age, and can last for several months. It is caused by cytomegalovirus (CMV) or Epstein-Barr virus (EBV), and usually results in prolongued cold and flu symptoms, swollen lymph nodes, and fatigue. In 1981 a group of researchers looked at CD4 and CD8 counts in ten consecutive patients with acute CMV mononucleosis, and compared their counts with those of ten healthy volunteers (Carney et al. 1981). The CD4 counts in people with mononucleosis were signifcantly reduced, with the healthy volunteers having 73% more CD4+ cells per ml than did people with mono, on average. The CD8 cells in people with mono were increased, and the combination of lower CD4 counts and elevated CD8 counts resulted in an inverted CD4/CD8 ratio in every patient. The average ratio was only 0.2, compared to the normal average of 1.7 found in controls. The CD4 counts were measured in nine of the ten patients, and the three with the lowest CD4 counts had 194, 202 , and 255 cells/mm3. The authors also found that the T-lymphocytes of people with mononucleosis responded poorly to antigens, showing depressed function. this paper was published three years before HIV was first claimed to be the cause of AIDS.

Five years later, a different set of researchers measured various lymphocyte subsets in acute EBV mononucleosis (Junker et al. 1986). They took 17 consecutive patients who had recently been diagnosed, gave them an immunization designed to activate their B lymphocytes, and then took samples of blood. The immunization makes this study different from any of the other studies to be examined here. They did not find a statistically significant lowering of CD4 counts, but they did find significantly lowered CD4/CD8 ratios, with the ratios falling below 1 as is reported to occur in people diagnosed HIV-positive. The authors conclude that "these studies demonstrate that infection with EBV affects both B and T lymphocytes and causes a broad based transient immune deficiency". The immune deficiency was transient but long lasting, with CD4/CD8 ratios gradually returning to normal over the course of 4 to 6 weeks.

2d) Low CD4 counts in sepsis

In 1986, a group of researchers from Osaka, Japan published a study where they examined various lymphocyte subsets in 9 consecutive patients admitted to the ICU with sepsis (Nishijima et al. 1986). They examined their blood at weekly intervals for four weeks. The CD4 counts in these patients were markedly reduced, with averages beginning below 500 and staying there for the entire 4 week study period. They also found T-cell function to be diminished, especially in patients who did not survive, although there was no significant difference in CD4 counts between those that died and those that survived. The authors did not provide individual CD4 counts.

2e) Low CD4 counts in pulmonary tuberculosis

Tuberculosis is a relatively common infection in people diagnosed HIV-positive, especially when compared to the general population. It is also relatively common in other people who are immunosuppressed, such as alcoholics, the homeless, intravenous drug users (IVDUs), and people who suffer from malnutrition. In 1985 a group of researchers in Indonesia examined the lymphocyte subsets in 26 patients newly diagnosed with pulmonary tuberculosis (TB) (Beck et al. 1985). They undertook the study because of a previous report of lowered CD4 counts in HIV-positive patients with TB in which the authors assumed that the lowered CD4 counts were due to HIV. They found that in HIV-negative TB patients CD4 counts were also significantly lowered, with an average of 748, compared to 1,043 in healthy controls. They also found significantly lowered CD4/CD8 ratios. Although the effects seen here were not as dramatic as in the studies reviewed previously, with only 5 of 26 patients having CD4 counts less than 500, the authors still felt their findings were highly significant to people diagnosed HIV-positive. Here are some of their comments:

In a study of AIDS, ... Vieira et al. stated that it was possible, but highly unlikely, that tuberculosis or its treatment could have altered the relative numbers of circulating lymphocytes bearing the markers CD4 ... and CD8, but they dismissed this possibility because of the severity of the altered CD4/CD8 ratio... We now report the relatively frequent occurence of moderate CD4 lymphopenia in patients with untreated but otherwise uncomplicated pulmonary TB." (Beck et al. 1985, page 50)

The authors also comment on some similar findings in leprosy, as well as in HIV-negative hemophiliacs:

Moderate reduction in the CD4/CD8 ratio has been reported in lepromatous leprosy, which reverts to normal under effective chemotherapy... It is tempting to speculate that these changes are analogous to those we now report in tuberculosis and that they are a consequence of ongoing immune response to the disease... Interestingly, comparable CD4 lymphopenia has been reported haemophiliacs treated with factor VIII, from a population apparently free from AIDS, and this change has been attributed to a reaction to transfusion of foreign proteins. (Beck et al. 1985, page 53).

The reports of the effects of factor VIII transfusions on CD4 counts have since been confirmed, as will be presented in section 3.

2f) Nearly all viruses interfere with lymphocyte function

In 1987 a summary article appeared in the Annual Review of Immunology entitled simply, "Viruses Perturb Lymphocyte Functions" (McChesney & Oldstone 1987). This article did not look at CD4 counts, but rather focused on the ability of CD4+ T-cells and other cells to proliferate when presented with an antigen. The authors reviewed evidence that a multitude of viruses interfere with the ability of CD4+ T-cells to proliferate. Following are some direct quotes from the text:

Viruses with every type of genomic nucleic acid, encompassing divergent replication strategies, are now known to infect lymphocytes. The list (Table 1) of viruses that infect lymphocytes is not comprehensive, but rather indicates representative viruses from different taxonomic groups. With few exceptions, immunologic dysfunction has been associated with the infections. (McChesney & Oldstone 1987, pages 280-281)

The viruses listed in Table 1, that infect human lymphocytes, are: Hapatitis B virus, Group C adenoviruses, Herpes simplex viruses, Cytomegalovirus, measles, mumps, respiratory syncytial virus (RSV), Vesicular stomatitis virus, Inflenza A, Parainfluenza, Rubella, Poliovirus, Lymphocytic choriomeningitis virus, and Human T-cell leukemia viruses I and II.

After a lengthy section focusing on the measles virus, they go on to discuss the class of viruses to which HIV belongs, retroviruses:

Retroviruses of murine (mouse), avian (bird), feline (cat), and human origin are immunosuppressive as well as oncogenic in their hosts. The evidence of depressed cellular and humoral immune responses... is independent of the transforming function of the virus... There is no interspecies restriction, i.e. both murine and feline retroviruses can suppress mouse and human lymphocyte proliferation in vitro. (McChesney & Oldstone 1987, page 287)

The authors go on to describe that it is not necessary for the entire retrovirus to be introduced, but only some of its proteins, so that the depressed response is apparently a passive one that does not require any action on the part of the viruses being discussed.

A partially purified 15-kd structural protein of a feline retrovirus inhibited the proliferation of feline lymphocytes... The inhibition was dose dependent and and occured when the protein was added as late as day 3 of a 4-day culture. In contrast, another structural protein, p27, was not inhibitory. (McChesney & Oldstone 1987, page 287)

Unfortunately, at least for the purposes of this paper, the authors do not discuss CD4 cells, specifically.

3) Low CD4 counts caused by injections of foreign proteins

3a) CD4 irregularities in hemophilia

Hemphiliacs were one of the original HIV risk groups. As mentioned above, hemophiliacs who are HIV-negative have been found to have lowered CD4 counts as well as lowered CD4/CD8 ratios, and it appears that this effect is caused by injections of factor VIII. Antonaci et al (1988) for example, found decreased CD4/CD8 ratios as well is impaired CD4 function in HIV-negative hemophiliacs, stating in their conclusion that "Our findings clearly indicate an impairment of immune function in hemophiliacs regardless of HIV infection" (page 318). Similarly, Madhok et al. (1986) found depressed cell-mediated immunity that was independent of HIV status. Their abstract contains the following comments:

There was no difference in skin response between patients positive and negative for the human immunodeficiency virus (HIV). In the whole group, and in seronegative patients (n = 17), there was an inverse relation between exposure to clotting factor and skin response. In seropositive patients (n = 12) no such association was apparent. This study shows that clotting factor concentrate impairs the cell mediated immune response to a new antigen in the absence of infection with HIV. (Madhok et al. 1986, page 978)

3b) CD4 irregularities caused by injected drugs

Intravenous drug users (IVDUs) are another group with a high risk of being diagnosed HIV-positive. In an article published in 1987 in the journal, AIDS, lymphocytes were found to be reduced in HIV-positive injection drug users as a direct function of how many injections they received (Des Jarlais et al. 1987). The authors comment in their abstract:

Continued drug injection was associated with the rate of CD4 cell loss... While it is not possible to distinguish the mechanism underlying the relationship between continued drug injection and CD4 cell loss, seropositive IV drug users should be warned that continued injections may lead to increased HIV-related immunosuppression. (Des Jarlais et al. 1987, page 105)

A similar finding in 1991, also published in the journal, AIDS, found that lymphocyte reactivity was much more significantly reduced in IVDUs who injected more frequently, regardless of whether or not they were HIV-positive (Mientjes et al. 1991). Although the CD4 cell function was impaired, no difference was found in CD4 counts due to frequent injecting. They did find that HIV-positive IVDUs had lower CD4 counts than did HIV-negative IVDUs, however. The T-cell reactivity was 40-50% lower in IVDUs who were injecting 3 times a day for the preceding several months when compared to a similar group who had not injected in the preceding months, regardless of their HIV status. The authors write: "We conclude that lymphocyte reactivity is depressed by frequent injecting in both HIV-negative and HIV-positive drug users" (Mientjes et al. 1991, page 35).

As far back as 1980, a report in the Journal of Immunology documented lowered T-lymphocytes in IVDUs from Georgia, Illinois, and Massachusetts (McDonough et al. 1980). The authors found that IVDUs in their study had about half to one third as many T-lymphocytes, expressed as a percentage, as control populations. Although they did not look specifically at CD4+ T-lymphocytes, it has been found that when total T-lymphocytes are reduced, CD4 counts are also normally reduced (Kotze 1998). They discuss previous findings of opiate receptor sites on T-lymphocytes, suggesting that the IV opiates were the cause of the lowered T-cells, but they also recognize other possible contributing factors:

Since most street heroin addiction involves polydrug use including chronic use of marijuana, barbiturates, hallucinogens, and other illicit substances, the hypothesis can be proposed that the depression of T-lymphocyte percentage was caused by another drug or combination of drugs, or by the effect of drug use on the addict's general physical health and nutrition, i.e., the addict milieu. (McDonough et al. 1980, page 2542)

The finding that a wide variety of physical and psychological stressors can lower CD4 counts supports this multifactorial argument, in which general health and nutrition can be significant contributing factors.

Finally, a review paper that was published in 1995 in the journal, Immunopharmacology, had an interesting discussion of the significance of this information for IVDUs diagnosed HIV-positive.

Among the unwarranted side effects of respiratory depression, constipation, and physical dependence are the immunosuppressive qualities, particularly those which affect cell-mediated immunity. The immunosuppressive characteristics of opioid narcotics (e.g., morphine) have recently come into focus with the advent of acquired immune deficiency syndrome (AIDS) and the putative causative agent, human immunodeficiency virus type 1 (HIV-1). Specifically, a vast reservoir of HIV-1-infected individuals exists among drug abusers. Moreover, experimental evidence would suggest narcotic opioids may increase viral load in infected individuals. (Carr et al 1995, page 59).

3c) CD4 Irregularities caused by in utero exposure to opiates

In 1987, a study found that infants exposed to intravenous drugs in utero also have decreased CD4/CD8 ratios and reduced CD4 function, even when they are HIV-negative (Culver et al 1987).

The CD4/CD8 ratio decreased with age in the drug-exposed infants compared with control infants (P less than 0.005). ... Our data demonstrate that infants of intravenous drug-using mothers have distinct immunologic differences at birth compared with non-drug-exposed infants and that these persist throughout the first year of life. The cause appears unrelated to intrauterine viral infection, suggesting a direct toxic effect of the drugs on fetal immunologic development. (Culver et al. 1987, page 230)

These results show that multifactorial causes of low CD4 counts probably apply to all age groups, including newborns. This is especially true in the United States and in Europe where most newborns who are HIV-positive are born to women who use intravenous drugs. In Africa, malnutrition and other infectious diseases are more likely to contribute, as will be discussed below.

4) Low CD4 counts caused by injuries and burns

Several studies over the years have looked at the effects of severe injuries or burns on CD4 counts. An early report appeared in 1982, in which the authors looked at the percentage of CD4 counts in 30 patients admitted to their hospital's burn center (Antonacci et al. 1982). They found that the severity of the burns was directly correlated with depressed CD4 percentages. Patients with greater than 25% of their body covered with 3rd degree burns had the lowest percentages on admission, 37%, as compared to normals who had 63%. They found a similar pattern with the CD4/CD8 ratio, but do not report on absolute CD4 counts.

In 1984, a group of researchers decided to look at lymphocyte subsets in patients with multiple trauma who had no burns (O'Mahoney et al. 1984). They examined the blood of 31 patients and compared their lymphocyte profile to ten normal controls. The CD4/CD8 ratio was significantly reduced and inverted, with an average of 0.96, compared with 1.82 in controls. They also found reduced lymphocyte proliferation/blastogenesis in response to antigen challenges. While their original report said they found no difference in absolute CD4 counts, they report in a postscript that they were mistaken in this regard: "in looking back now at the data, we feel the CD4 population did change relative to the CD8 population because of an absolute decrease in the number of CD4 cells" (O'Mahoney et al. 1984, page 875).

In 1985, a study was published that looked at two groups of patients with severe injuries, a group of 25 patients with burns, and a group of 21 patients with non-thermal injuries. Both groups had severely lowered CD4 percentages, which persisted until 50 days post-injury when the study was concluded. They also found that people with lower CD4 percentages were more likely to develop sepsis. Here are some of the author's comments:

The most important abnormality appears to be a reduction in CD4 positive cells in burn patients... A change in the ratio of CD4 to CD8 positive cells soon after injury is due to a reduction in CD4 positive cells, not an increase in CD8 positive cells (page 584).

We believe that the more important abnormality in the patients studied is a reduction in T-cell help - both in terms of the number of circulating CD4 positive cells and and a reduction in interleukin 2 production seen both in burn and non-thermal injury patients. Interleukin 2 is produced by T-cells, especially CD4 positive cells, and promotes their growth and stimulates clonal expansion of T-cell subsets: it is thus crucial in the response to foreign antigen. (O'Mahoney et al. 1984, page 585).

The final study to be reviewed is also quite old, from 1986, and looked at 20 consecutive patients who had emergency surgery due to major trauma (Polk et al. 1986). This was the only study of trauma victims where absolute numbers of CD4 cells are given, which makes it more significant from the perspective of this paper. Figure 7 on page 289 shows that 6 of 20 (30%)patients had CD4 counts below 200 cells/mm3, and 13 of 20 (65%) had counts below 500. The authors state simply: "Total T-cells represent what is interpreted as a normal and common response to injury... All patients had low total lymphocyte counts on admission and exhibited a furthur decline on day 3" (Polk et al. 1986, page 287). 10 of the patients also had major infections, and three had minor infections, which may have also contributed to their extremely low CD4 counts. This paper is distinctive in that it attempts to explain a mechanism for the lowered CD4 counts, citing a study supporting the hypothesis that increased cortisol levels are responsible for the decline, and that increased cortisol is also a normal response to injury. The argument that cortisol plays a key role in lowered CD4 counts will be encountered again in the section on psychological stress.

5) Low CD4 counts in normal human pregnancy

Several studies have been published on CD4 counts during normal pregnancy. Most recently, Burns et al.published a study in 1996, where they attempted to control for potentially confounding factors like the increased blood volume that normally occurs in pregnancy. They used CD4 percentages because of this variable, and determined that "Our CD4 cell findings for HIV-negative women are consistent with the majority of prior studies, which demonstrate a decline in CD4 levels during normal pregnancy" (Burns et al. 1996, page 1465). They also found that HIV-positive women had a more severe decline which did not correct post-partum as it did in HIV-negative women, although they fail to take into account other factors that can cause lowered CD4 counts. These include any infections that the women may have experienced, the traumatic effects of C-sections which are normally performed on HIV-positive women to prevent neonatal transmission, or the potentially severe psychological stress of worrying if their baby will also be HIV-positive, which can take up to 18 months to determine.

In 1989 a study was published of normal pregnancy which found reduced CD4 percentages in the 1st and 2nd trimester, as well as reduced CD4/CD8 ratios in the 2nd trimester (Castilla et al. 1989). They comment on previous studies looking at a variety of lymphocyte changes during pregnancy, stating simply, "In these studies, variation in the number and proportion of CD4+ lymphocytes is the alteration most frequently reported" (Castilla et al. 1989, page 104). They also claim that "we have accounted for all the presently known factors that can alter the concentrations of T-cell subsets in blood" (Castilla et al. 1989, page 104), but in fact they did not consider any of the factors described in this paper, such as infections, trauma, overexercising, normal daily variation, or psychological stress. This demonstrates that even clinicians and researchers doing studies that focus specifically on CD4 levels are often unaware of how many different conditions cause low CD4 counts.

The final study to be reviewed here is an early one from 1982 (Sridama et al. 1982). These researchers found reduced absolute CD4 counts, as well as reduced percentages of CD4+ T-cells in 76 women with normal pregnancies. By the third trimester, the pregnant women had an average of only 543 + 169 CD4+ T-cells, compared to 1073 + 441 in non-pregnant women who served as controls. Both the absolute numbers and the percentages stayed low until several months post-partum, and similar results were obtained for the CD4/CD8 ratio. This is the only study found of normal pregnancy that provides data on absolute CD4 counts, and the average of 543, with a standard deviation of 169, means that a relatively large percentage of these women had levels lower than 500, the point at which antiretroviral medications would be started in someone diagnosed HIV-positive.

6) Reduced CD4 counts from overexercising

Only one study will be discussed here, which was published in 1992 (Verde et al. 1992). In a controlled trial, ten athletes were asked to over-train for three weeks. Blood samples were taken immediately before starting, at the end of the three weeks, and again three weeks after returning to normal. The researchers found steady declines in the percentage of CD4+ T-cells, with the lowest amount occuring 3 weeks after returning to a normal exercise schedule. The authors also found reductions in the CD4/CD8 ratio, although these had normalized by the 3 week endpoint. Finally, the authors also checked levels before and 5 minutes after acute exercise, and again found reductions in CD4 percentages and in CD4/CD8 ratios, although these normalized by 30 minutes post-exercise. It is interesting that a stress as simple as overexercising for three weeks could cause lowered CD4 counts, and that they did not correct for at least three more weeks after returning to a normal exercise schedule.

7) Low CD4 counts in malnutrition

A number of studies have looked at the immunosuppression that results from malnutrition. Like the other conditions covered in this paper, malnutrition causes severe immunodeficiency with depletion of CD4+ T-cells and reduction of cell mediated immunity. One of the most recent studies is from India, where malnutrition is extremely common (Hegde et al. 1999). The authors found that reduced CD4 counts were a natural physiological effect of malnutrition, and comment that both HIV and malnutrition lead to a state of anergy with failure of cell-mediated immunity. They also point out that HIV usually occurs in conjunction with several other stressors of the immune system: "micronutrient abnormalities, concomitant infections, and genetic factors are some of the compounding co-factors which furthur contribute to the deterioration of immune functions in AIDS patients" (Hegde et al. 1999, page 318).

A review paper from the Journal of Nutrition in 1996 also compares malnutrition and AIDS, saying that "Protein/energy malnutrition or deficiencies of single nutrients that assist in nucleic acid metabolism generally lead to atrophy of lymphoid tissues and dysfunctions of cell mediated immunity" (Beisel 1996, page 2611S). The author comments on a syndrome of immunosuppression caused by malnutrition which is called "NAIDS", and states that it often occurs in people diagnosed HIV-positive:

Immunological dysfunctions associated with malnutrition have been termed Nutritionally Acquired Immune Deficiency Syndromes (NAIDS). Infants and small children are at great risk because they possess only immature, inexperienced immune systems and very small protein reserves. The combination of NAIDS and common childhood infections is the leading cause of human mortality. NAIDS can generally be corrected by appropriate nutritional rehabilitation, but from a viewpoint highly important to this Workshop, AIDS and NAIDS are intensely synergistic... Aggressive nutritional support for children with HIV infections could delay, or lessen, the development of NAIDS and avoidance of NAIDS would improve both quality and length of life. (Beisel 1996, page 2611S)

Later in the paper they describe some of the immunological changes and clinical courses often seen in malnutrition, which sound very similar to AIDS.

Generalized, protein energy malnutrition causes widespread atrophy of lymphoid tissues, especially in children. The thymus, spleen, tonsils, and lymph nodes are all affected, with evidence of atrophy being greatest in T-lymphocyte areas of these tissues. ...

Malnutrition, in turn, leads to a variety of immune system dysfunctions, ... which allow infectious diseases to fluorish. These closely linked events can initiate a "downhill spiral" or a "vicious cycle" that leads inexorably to death.

Protein energy malnutrition causes a marked repression of cell-mediated immunity and the function of T-lymphocytes. Malnourished children show anergy with loss of delayed dermal hypersensitivity reactions and a decrease or reversal of the CD4/CD8 cell ratio... In contrast, B-lymphocyte numbers and functions appear to be maintained. While existing antibody production is conserved or even increased during malnutrition, antibody responses and antibody affinity are impaired. (Beisel 1996, page 2612S)

The "downhill spiral" of opportunistic infections that "lead inexorably to death" is particularly reminiscent of a description of AIDS. Beisel also reviews similar effects of deficiencies of specific nutrients, such as vitamin A and zinc:

Deficiencies of single essential nutrients with important roles in nucleic acid synthesis and metabolism appear to cause derangements in immunological functions that are quite similar to those seen in protein energy malnutrition ... Both vitamin A and zinc deficiencies are characterized by lymphoid tissue atrophy and depressed cellular immunity ... (Beisel 1996, page 2613S)

To provide an idea of how prevalent the problem of malnutrition is worldwide, he points out that the combination of malnutrition induced immunosuppression and childhood infections "is the leading cause of human mortality, producing more than 10 million deaths per year (i.e. over 25,000 deaths per day)" (Beisel 1996, page 2614S).

Another review paper published one year later, in 1997, made similar arguments about the significance of malnutrition in impairing immunity (Chandra 1997). This is the only paper found that gives percentages of CD4 cells, although absolute CD4 counts are not provided. Figure 3 on page 462S shows that the percentage of CD4+ T-cells in normal well-nourished children is about 45%, while the percentage in malnourished children is only 25%. Chandra describes the immune system changes seen in malnutrition:

Nutrition is a critical determinant of immune responses and malnutrition the most common cause of immunodeficiency worldwide. Work done in the past 25 years has confirmed that impaired immunity is a critical adjunct factor in malnutrition-associated infections. ... Lymphoid atrophy is a dramatic feature of protein energy malnutrition. ... Delayed hypersensitivity cutaneous responses are markedly depressed. It is not uncommon to have complete anergy to a battery of different antigens. These changes are observed in moderate deficiencies as well. The skin reactions are restored after appropriate nutritional therapy for weeks or months. ... the proportion of helper T-lymphocytes (CD4+ T-cells) is markedly decreased, and the ratio of CD4 to CD8 cells is significantly lower than in well-nourished control subjects. (Chandra 1997, page 460S-461S)

From this review it is seen that not only are the CD4 percentages markedly reduced (from 45% to 25%), but that it takes "weeks or months" of nutritional therapy for the effects of malnourishment to revert to normal.

The final article to be examined is also a review (Harbige 1996). This paper mentions similar findings to the ones already discussed, including lowered CD4+ lymphocytes, decreased T-cell function, and anergy. It also mentions the increase in antibody levels which is also seen in people diagnosed HIV-positive, specifically serum IgG, IgM, IgA, and IgD. In contrast to serum IgA, however, secretory IgA is diminished. The main addition that this paper provides which others did not is the mention of specific infections that are particularly common in people who are malnourished:

Among the many infectious organisms commonly associated with protein energy malnutrition are Paramyxovirus (Measles), Rotaviruses, Mycobacterium tuberculosis, E-coli, Shigella, E-histolytica, and Pneumocystis carinii. (Harbige 1996, page 289)

Two of these organisms, M. tuberculosis and Pneumocyctis Carinii, result in the diagnosis of AIDS when they occur in someone who has already been diagnosed HIV-positive. Pneumocystis carinii pneumonia, or "PCP", is perhaps the single infection most commonly associated with AIDS.

This information concerning malnutrition-induced immunodeficiency and opportunistic infections is obviously significant for Africa, where malnourishement is common and where HIV is also thought to be the most prevalent, but it also may be very significant for people in the United States and Europe. Several articles point to malnourishment as a very common problem in AIDS due to decreased nutrient intake or malabsorption (Babameto & Kotler 1997, Keusch & Thea 1993). These can be caused by infections of the oral cavity and gastrointestinal tract, which are quite common in people diagnosed with AIDS. Antiretroviral medications, which cause diarrhea and/or vomiting in well over half of the people who take them, also have the potential to interfere significantly with nutrient intake, especially when combined with other factors. In addition, decreased appetite is one of the standard symptoms of depression, which is common in people diagnosed HIV-positive. Finally, infections of any type put a physical stress on the system which results in loss of weight. This is because people break down their own tissues to use as fuel, resulting in increased nutrient requirements. It could easily be argued that food and financial independence are solutions that should be given a much higher priority when offering aid to poorer nations, and these studies suggest that food and social supports should be a higher priority for HIV and AIDS programs in wealthy nations, as well.

8) Daily variation of CD4 counts

Only one study concerning the daily, or diurnal, variation in CD4 counts will be reviewed here (Malone et al. 1990). The authors compared the diurnal variation in HIV-positive and HIV-negative people, finding a significant variance in both. They found that greater variations occurred in HIV-negative people, but that both groups followed a pattern that coincides with known daily fluctuations of cortisol, with minimum CD4 levels occuring between 8:00 and 10:00 a.m., and maximums occuring at around 10:00 p.m.. Cortisol has a daily variation with maximums at about 8:00 a.m., and, as will be reviewed later in this paper, cortisol also causes low CD4 and total T-lymphocyte counts. People with lower baseline CD4 counts had much less diurnal variation. This pattern is often seen in people under chronic stress where the normal variation of cortisol secretion is severely flattened. HIV-negative people had an average variation of 506 cells/mm3 each day, while HIV-positive people had only about 60 cells/mm3 of variation. The authors caution that even this blunted variation is significant, however, stating "3 of 12 HIV-positive patients had CD4+ cell counts below 200 cells/mm3 in the morning but had greater than 200 cells/mm3 in the afternoon" (Malone et al. 1990, page 150). In other words, in the morning they would be diagnosed with AIDS, but if their blood was checked in the afternoon they would just be HIV-positive, albeit with a relatively low T-cell count. They found similar results for total lymphocyte counts, but CD4/CD8 ratios did not have statistically significant changes. The authors conclude that blood draws for CD4 counts should always be done at the same time of day, but they do not comment on relations between the diurnal cycle they observed and the diurnal variation in cortisol.

9) Changes in CD4 counts and lymphocyte function due to psychological stress and social isolation

A large number of studies have looked at the effects of stress on the immune system, and several reviews have been published on this topic (Bonneau 1993, Castle 1995, Herbert 1993, Kennedy 1988, Kiecolt-Glaser 1984, 1991, 1992 Laudenslager 1983, Pariante 1997, Stefanski 1998). These studies have looked at people under chronic stress, such as people suffering from depression, people who were recently divorced or separated, college students during exams, and people who are the primary caregivers of demented family members. There are also a number of studies of animals under stress. Stress causes a state of immunodeficiency characterized by a reduction of the number of T-lymphocytes, with special targeting of CD4, helper T cells. There is also a reduced CD4/CD8 ratio, with a relative increase in CD8, suppressor/cytotoxic T cells. Unfortunately for the purposes of this paper, the vast majority of studies look at lymphocyte function and total T-cell counts. The few studies that have looked at CD4 cells used percentages(Kiecolt-Glaser et al. 1992).

A group of researchers led by Robert Sapolsky has done a great deal of work observing the effects of psychological and social stress on baboons and other primates, with most of their work focusing on the neurotoxicity that is caused by stress, with dementia and loss of neurons in the hippocampus (Sapolsky 1990, 1996). In one study, however, they measured total lymphocyte counts and cortisol levels in a group of baboons that were invaded by a highly aggressive young male baboon, whom they named Hobbs (Alberts et al. 1992). Hobbs was particularly threatening to females in the group, and was apparently attempting to use fear, physical intimidation, and abuse to increase his chances of successful mating. Cortisol levels in the group nearly doubled after Hobbs joined the group, with a slightly greater increase among females. T-lymphocytes plummeted in the group, from a pre-Hobbs level of 67 per 10,000 red blood cells to a level of about 39, a drop of 42%. When looking at only the levels in baboons who were victims of Hobbs' aggression, the levels fell even more steeply, to only 29 per 10,000 RBC's, a drop of 55%. Interestingly, Hobbs, himself had the lowest number of lymphocytes in the entire group, and the highest cortisol level, suggesting that his behavior may have been taking an even greater toll on his system than it did on the victims of his aggression. Field conditions prevented them from determining the number of lymphocytes per microliter of blood, or from specifically measuring CD4 cells, and the authors comment on their use of lymphocyte counts instead of more sophisticated methods:

Whereas most studies of the effects of stress upon immunity examine functional indices of immune competence (e.g. mitogen stimulation tests, antibody generation, cytokine responsiveness), our field conditions limited us to this rather crude quantitative measure of numbers of cells. (Alberts 1992 page 174)

It is interesting that these researchers consider T-cell counting to be a crude measure of immune competency. Although the clinicians in this study could not report on CD4 counts, low total lymphocyte counts are associated with low CD4 counts (Kotze 1998), so the findings of this study are likely to indicate that CD4 counts are also lowered..

A review from as far back as 1988 examined how the immune system was affected by stress, with the following comments regarding CD4 helper T-cells (Kennedy et al. 1988):

Data are given which document immunosuppressive effects of commonplace, short-term stressors, as well as more prolonged stressors, such as marital disruption and caregiving for a relative with Alzheimer's disease. Immune changes included both quantitative and qualitative changes in immune cells, including changes in herpes virus latency, decreases in the percentages of T-helper lymphocytes and decreases in the numbers and function of natural killer cells. These effects occurred independently of changes in nutrition. Psychological variables, including loneliness, attachment and depression were related to the immune changes. The data are discussed in a framework in which quality interpersonal relationships may serve to attenuate the adverse immunological changes associated with psychological distress, and may have consequences for disease susceptibility and health.' (Kennedy et al. 1988, page 77).

Another review, published in 1993, performed a meta-analysis of all studies that looked at psychological stress and the immune system (Herbert & Cohen 1993). In their discussion they mention their findings regarding CD4, helper T-cells:

In terms of cell numbers, stress is reliably associated with a ... lower number of circulating B cells, helper cells, cytotoxic cells, and large granular lymphocytes. Stress is also reliably associated with a lower percent of lymphocytes that are T cells, helper T cells, and cytotoxic T-cells. (Herbert & Cohen 1993, page 373)

The last review to be discussed here looked at short-term stressor effects and made similar comments to the two reviews above, again focusing on CD4 percentages instead of absolute CD4 counts:

The immunological changes observed following short-term stressors are very similar to those that have been described following epinephrine injections: increased percentages of natural killer cells, decreased blastogenesis in response to mitogens (decreased lymphocyte function), and decreased percentages of CD4 cells. Total T cells and monocytes did not change. (Kiecolt-Glaser et al. 1992, page 680)

This quote mentions epinephrine injections, but cortisol injections also produce similar effects on the immune system. Secretion of these hormones is the most commonly proposed mechanism for the immunosuppression that occurs during states of acute or chronic psychological stress. One of the major changes during times of stress is an outpouring of the hormones epinephrine and cortisol, which lead to a dramatic reduction in the number of T-lymphocytes. The strength of the correlation between decrease in T-cells and excess cortisol is so strong that low T-cells is one of the diagnostic criteria for identifying excess cortisol. Here are some quotes on this topic from a basic textbook of physiology (Guyton 1996).

"Almost any type of physical or mental stress can lead within minutes to greatly enhanced secretion of ACTH and consequently cortisol as well, often increasing cortisol secretion as much as 20-fold" (Guyton 1996, p.966).

"Cortisol suppresses the immune system, causing lymphocyte production to decrease markedly. The T lymphocytes are especially suppressed." (Guyton 1996, p.964)

"Cortisol decreases the number of eosinophils and lymphocytes in the blood; this effect begins within a few minutes of injection of cortisol and becomes marked within a few hours. Indeed, a finding of lymphocytopenia or eosinopenia is an important diagnostic criterion for overproduction of cortisol by the adrenal gland. Likewise, the administration of large doses of cortisol causes significant atrophy of all the lymphoid tissue throughout the body... This occasionally can lead to fulminating infection and death from diseases that would otherwise not be lethal, such as fulminating tuberculosis in a person whose disease had previously been arrested" (Guyton 1996, p.965).

It is interesting that this description of 'fulminating infection and death from diseases that would otherwise not be lethal' sounds very similar to a description of AIDS. There is a disease which is characterized by long-term hypersecretion of cortisol, called Cushing's syndrome or Cushing's disease. Cecil Essentials of Medicine describes the physical manifestations of Cushing's disease, many of which are also common in AIDS:

Regardless of the etiology, hypercorticolism results in central obesity, carbohydrate intolerance, muscle wasting, and osteoporosis. Obesity is centripetal, manifested typically by a "buffalo hump", increased supraclavicular fat pads, and moon facies... Depression occurs often, and, rarely, patients may be frankly psychotic. (Andreoli et al. 1993)

Cushing's disease also causes immunodeficiency (Britton et al. 1975) and loss of cortical neurons (Starkman et al. 1992).

It is especially interesting that the redistribution of fat described here is a common side effect seen in HIV-positive patients after long-term protease inhibitor use, with the same "buffalo hump" and central obesity ("crixivan belly"). Also, early osteoporosis has been recently found to be another common adverse effect of these medications. Multiple studies have found that people diagnosed HIV positive have chronically elevated cortisol levels, suggesting that the low CD4 T-cells in people diagnosed with AIDS could be at least partly caused by elevated cortisol (Azar 1993, Christeff 1988, 1992, Coodley 1994, Lewi 1995, Lortholary 1996, Membreno 1987, Norbiato 1996, Norbiato 1997, Verges 1989). It is important to note, however, that chronic stress can induce immune suppression even when cortisol and epinephrine are not elevated (Bonneau 1993, Keller 1983), so that the mechanisms by which stress affects health and immunity are not completely understood.

In 1998, a group of researchers put these hypotheses to the test by checking CD4 counts and cortisol levels in people who were randomly assigned either to a bereavement support group intervention or to a wait-list control (Goodkin et al. 1998). The intervention consisted of 10 weekly support group meetings, and blood samples continued to be taken periodically for a total of 6 months. Some of the group members were HIV-positive, and the authors stratified their data according to HIV status. They found that CD4 counts were increased in people receiving the support group intervention as compared to controls, and that these increases correlated with reduced levels of the stress hormone cortisol. Here is their description of the results:

In HIV-negative intervention subjects, the CD4 cell count increased 112 cells/mm3, while that in HIV-negative control subjects decreased 88 cells/mm3, for a difference of 200 cells/mm3 between treatment and control groups. In treated HIV-positive individuals, the CD4 cell count was stable, within laboratory error over the entire six months. However, that in HIV-positive controls decreased 61 cells/mm3. Both (statistical tests) demonstrated a statistically significant intervention effect on the CD4 cell count. (Goodkin et al. 1998, page 387)

Results like these may help to explain why socially isolated people, when compared to people with high levels of social support, have been found in over eight studies to have between double and triple the death rates from all causes (Berkman 1979, House 1988, Ornish 1997). A recent study found that people diagnosed HIV positive were two to three times more likely to 'progress to AIDS' if they were socially isolated and under high levels of stress (Leserman et al. 1999). Here is a brief quote from the abstract of their paper:

Faster progression to AIDS was associated with more cumulative stressful life events (p<0.002), more cumulative depressive symptoms (p<0.008), and less cumulative social support (p<0.0002). ... At 5.5 years, the probability of getting AIDS was about two to three times as high on those above the median on stress or below the median on social support. ... (Leserman et al., page 397)

This study was not able to assess the impact of the stress of living with the diagnosis, HIV-positive, nor can any study that is ethically designed assess this impact. It is not unreasonable, however, to infer that the stress of the diagnosis is a contributor to mortality.

In addition to the cortisol hypothesis, another mechanism was presented in a paper in the journal, Medical Hypothesis, which also argues for a multifactorial model of AIDS (Shallenberger 1998). The author argues that the immune system becomes overbalanced towards antibody-mediated immunity (AMI) when it is chronically stressed. When this happens the cytokines released by this arm of the immune system (interleukins 4 and 10) naturally suppress the other arm, called cell-mediated immunity (CMI). CMI uses CD4+ cells in abundance, and when it is suppressed the CD4 count will drop. If the AMI dominance is maintained long enough it can become pathological and be very difficult to reverse, eventually leading to failure of both AMI and CMI. His arguments are supported by the fact that people who are diagnosed HIV-positive invariably have extremely high levels of antibodies, even when their CD4 counts have dropped significantly. Shallenberger carefully documents that evidence of this phenomenon occurs in all the risk groups for HIV, whether or not they are HIV positive, including hempheliacs, male homosexuals, IVDUs, and transfusion recipients. This is another possible mechanism, although it could still be mediated, at least in part, by excess cortisol secretion, which the author does not discuss.


This review is extremely limited in scope, and many of the studies presented here use different measures of immune function, making it difficult to perform accurate comparisons. Nevertheless, it is remarkable that so many different conditions are associated with profoundly reduced CD4 counts, as well as reduced CD4 percentages, reduced measures of lymphocyte function, and reduced CD4/CD8 ratios. The fact that HIV-negative people with many common conditions like mononucleosis, pregnancy, and pneumonia can have levels below those needed to diagnose AIDS suggests the common use of CD4 counts to make diagnostic and treatment decisions should be carefully reappraised, especially since most clinicians are apparently unaware of most of these possible influences. It is also apparent that the syndrome of AIDS, "acquired immunodeficiency syndrome", with extremely low CD4 counts and severe or fatal infections, is also fairly common in people diagnosed HIV-negative, and is likely to be present in at least 40% of people admitted to intensive care units with severe acute or chronic infections. In people who die of the infections they are admitted for this percentage is likely to be even higher.

It would be helpful to see studies of CD4 counts in even more common illnesses like influenza, and to see studies that try to determine when the CD4 counts begin to fall. If the CD4 counts were low before experiencing the conditions presented, the CD4 count could have caused the condition. The low counts in burn and trauma victims, however, argue in favor of the hypothesis that the conditions themselves caused the low CD4 counts, as it is difficult to argue that such a high percentage of people had low CD4 counts before the trauma was experienced. The studies reviewed here show that the CD4 counts can stay low for weeks or months, and this effect would be magnified if many factors were present at once or if several conditions occurred in sequential order, so repeated findings of low CD4 counts over time could also be a common finding. Studies looking at the mechanisms could also be helpful. The hypotheses presented in some of the articles reviewed above regarding mechanisms include increased cortisol, antibody-mediated immune dominance, and malnutrition. It is possible that several of these could operate simultaneously, as well, or that they could occur sequentially in a cascade.

Low CD4 counts are associated with so many different physical and psychological stressors that it is very possible that HIV is just one of many factors causing immunosuppression in people diagnosed HIV-positive. Following is a brief list of factors commonly present in "high-risk" groups that could explain or contribute to a low CD4 count:

  1. In Africa, malnutrition, endemic infectious diseases, and social ostracism could all be equally strong factors in causing an acquired immunodeficiency.
  2. In the United States and Europe, AIDS is still primarily confined to the original risk groups, male homosexuals, IV drug users, and hemophiliacs, all of whom regularly experience many of the conditions described:
    1. Male homosexuals suffer from societal rejection which causes psychological stress and social isolation. When AIDS first appeared, several events had made the social isolation of male homosexuals painfully clear; a successful campaign to repeal gay rights in Miami Dade County was being led by Anita Bryant, and the first elected official in the United States who was openly gay, Harvey Milk, was assassinated.
    2. The small subset of gay men in whom AIDS first appeared were engaging in a type of party atmosphere which involved multiple partners, late nights, and the regular use of alcohol and recreational drugs, possibly as a way to cope with the societal rejection they were experiencing. This lifestyle is probably what led in many people to recurrent and chronic sexually transmitted diseases.
    3. IV drug users live in conditions of psychological stress and social isolation, and also often suffer from malnutrition. They have always had high rates of infectious diseases including cellulitis, tuberculosis, pneumonia, and non-healing ulcers.
    4. Hemophiliacs need regular transfusions of factor VIII. The quality of the Factor VIII has steadily improved over the years, as has their health and life expectancy, but in spite of this they still have chronic health problems and their life expectancy is still greatly reduced when compared to the normal population.
  3. The very diagnoses, HIV-positive, carries a substantial burden of psychological stress and social isolation, which is made even worse when the CD4 count is found to be reduced, or when "full blown" AIDS is diagnosed.

All of these factors, from infections to psychological stress, could combine in causing immunosuppression, and many of them could be much more easily treated than current methods of treating HIV, which rely on long-term use of medications with a number of serious adverse effects.

Finally, the results described here cast doubt on the original claims that HIV specifically targets CD4+ T-cells. It could be instead that many or all of the conditions reviewed here operate under the same mechanism. The search for a new infectious agent began around 1979, when clinicians found extremely low CD4 counts in a few young male homosexuals who were dying of multiple infections. The ensuing international scientific search resulted in the Robert Gallo' claim that HIV was the cause, partly because it infected CD4+ lymphocytes (Gallo et al. 1984). Based on the results reviewed in this paper, however, it is very possible that the low CD4 counts in those early cases were simply a result of the opportunistic infections that were present, not because of any new agent that targeted CD4+ T-cells. This argument is strengthened by the continued difficulty in determining a mechanism for how HIV destroys CD4+ T-cells. The mechanisms originally proposed by Gallo have had to be abandoned, and new hypotheses have also had several major revisions over the years. A conference in 1997 determined that the cause was still unknown (Balter 1997). A quote from the article describing this conference, which appeared in the journal, Science, follows:

'It might be said that AIDS researchers know the virus that causes the disease, HIV, inside and out. They have isolated its proteins, sequenced its genome, and identified the receptors it uses to dock onto the CD4 T lymphocytes that are the viruses primary target. Yet the central mystery of AIDS remains unresolved: How does the virus cause the severe loss of CD4 T-cells, which wrecks the immune system, that is the hallmark of the disease?' (Balter 1997, page1399)

Another argument that appears to support the claim that HIV specificically targets CD4+ cells is that when anti-HIV medications are given, the CD4 count rises. This could simply be due to decreased infections, however, an argument that is bolstered by recent evidence that protease inibitors specifically inhibit Pneumocystis carinii and Candida albicans, two of the most common infections found in people diagnosed HIV-positive (Cassone 1999, Atzori 2000). This finding could also explain at least part of the beneficial effect that these medications are thought to have on people's health, and it is possible that they may inhibit other microbes as well. These medications also relieve some of the psychological stress that is created by the diagnosis, HIV-positive, which could also allow the CD4 count to rise. Once a rise in the CD4 count is seen, the relief of psychological stress would be strengthened. If other stressors that cause low CD4 counts are still present, however, these beneficial effects would be limited.

About 5-15% of people who are diagnosed HIV-positive do not go on to show any immunological abnormalities at all, even after ten or more years (Learmont et al. 1992, Walton 1999). In addition, only about 50% of people diagnosed HIV-positive will be diagnosed with AIDS in the first ten years after their diagnosis, a period which has been called the latent phase of the virus. Perhaps, by focusing on AIDS as a multifactorial illness, this latent phase can be extended indefinitely in more and more people. Furthur research that focuses on some of the many factors reviewed in this paper may also help reveal why these "long-term nonprogressors" appear to stay healthy in spite of being diagnosed HIV-positive, and may help increase the percentage of people who succeed in doing so.


Alberts SC, Sapolsky RM, Altmann J (1992). Behavioral, endocrine and immunological correlates of immigration by an aggressive male into a natural primate group. Hormones and Behavior 26; 167-178.

Andreoli TE et al. (1993). Cecil essentials of medicine. W.B. Saunders; Philadelphia.

Antonaci S, Jirillo E, Stasi D, De Mitrio V, La Via MF, Bonomo L (1988). Immunoresponsiveness in hemophilia: lymphocyte- and phagocyte-mediated functions. Diagn Clin Immunol;5(6):318-25

Antonacci AC, Good RA, & Gupta S (1982). T-cell subpopulations following thermal injury. Surg Gynecol Obstet; 155(1); 1-8.

Atzori (2000). In Vitro activity of HIV protease inhibitors against Pneumocystis carinii. J Infect Dis; 181; 1629-1634.

Azar ST, Melby JC (1993). Hypothalamic-pituitary-adrenal function in non-AIDS patients with advanced HIV infection. Am J Med Sci May;305(5):321-5.

Babameto G & Kotler DP (1997). Malnutrition in HIV infection. GI Clin North America: 26(2): 393-413.

Balter M (1997, November 21). How does HIV overcome the body's T-cell bodyguards? Science 278: 1399-1400.

Beck JS, Potts RC, Kardjito T, and Grange JM (1985). T4 lymphopenia in patients with active pulmonary tuberculosis. Clin Exp Immunol, Volume 60, 49-54.

Beisel WR (1996, october). Nutrition in pediatric HIV infection: setting the research agenda. Nutrition and immune function: overview. J Nutr;126(10 Suppl):2611S-2615S

Berkman L & Syme S (1979). Social networks, host resistance, and mortality: a nine year follow up study of alameda county residents. Am J Epidemiol; 109(2): 186-203.

Blatt SP, Lucey CR, Butzin CA et al. (1991). Total lymphocyte count as a predictor of absolute CD4+ percentage in HIV infected persons. JAMA 269; 622-626.

Bonneau RH, Sheridan JF, Feng N, Glaser R (1993). Stress-induced modulation of the primary cellular immune response is mediated by both adrenal-dependent and adrenal independent mechanisms. Journal of Neuroimmunology; 42; 167-176.

Britton S, Thoren M, Sjoberg HE (December 20, 1975). The immunological hazard of Cushing's syndrome. British Medical Journal 4; 678-680.

Burns DN, Nourjah P, Minkoff H, et al. (1996). Changes in CD4 and CD8 cell levels during pregnancy and post partum in women seropositive and seronegative for HIV-1. Am J Obstet Gyn; 174(5); 1461-1468.

Carney WP, Rubin RH, Hoffman RA, et al. (1981). Analysis of T lymphocyte subsets in CMV mononucleosis. The Journal of Immunology 126(6); 2114-2116.

Carr DJJ, Serou M (1995, November). Exogenous and endogenous opioids as biological response modifiers. Immunopharmacology; 31(1): 59-71

Cassone (1999). In vitro and in vivo anticandidal activity of HIV protease inhibitors. J Infect Dis; 180; 448-453.

Castilla JA, Rueda R, Vargas L, et al. (1989). Decreased levels of circulating CD4+ T lymphocytes during normal human pregnancy. J Reprod Immunol; 15; 103-111.

Castle S, Wilkins S, Heck E, Tanzy K, Fahey J (1995, September). Depression in caregivers of demented patients is associated with altered immunity: impaired proliferative capacity, increased CD8+, and a decline in lymphocytes with surface signal transduction molecules (CD38+) and a cytotoxicity marker (CD56+ CD8+). Clin Exp Immunol;101(3):487-93

CDC (1999). HIV/AIDS Surveillance Report. Centers for Disease Control, Atlanta, GA.

Chandra RK (1997, August). Nutrition and the immune system: an introduction. Am J Clin Nutr; 66(2) :460S-463S

Chirenda J (1999). Low CD4 count in HIV-negative malaria cases, and normal CD4 count in HIV-positive and malaria negative patients. Cent Afr J Med; Volume 45(9): page 248.

Christeff N, Gharakhanian S, Thobie N et al. (1992). Evidence for changes in adrenal and testicular steroids during HIV infection. J Acquired Imm Def Syn; 5: 841-846.

Coodley GO, Loveless MO, Nelson HD et al. (1994). Endocrine function in the HIV wasting syndrome. J Acquired Imm Def Syn; 7: 46-51.

Culver KW, Ammann AJ, Partridge JC, Wong DF, Wara DW, Cowan MJ (1987, August). Lymphocyte abnormalities in infants born to drug-abusing mothers. J Pediatr;111(2):230-5.

Des Jarlais DC, Friedman SR, Marmor M et al. (1987, July). Development of AIDS, HIV seroconversion, and potential cofactors for CD4 cell loss in a cohort of intravenous drug users. AIDS 1(2): 105-111.

Feeney C, Bryzman S, Kong L, Brazil H, Deutsch R, Fritz LC (1995, Oct). T-lymphocyte subsets in acute illness. Crit Care Med; 23(10):1680-5.

Fox CH (1996). The pathogenesis of HIV-disease. J Nutr; 126(10 Suppl): 2608S.

Gallo RC, Salahuddin SZ, Popovic M, et al (1984). Frequent Detection and Isolation of Cytopathic Retro-viruses (HTLV-III) from Patients with AIDS and at Risk for AIDS. Science ; 224:500-502.

Garrett L (2001). Change in Guidelines for HIV; U.S. officials to tout new treatment policy. Newsday (New York, NY), January 17, 2001, Wednesday, page A22.

Goldman (2000). Cecil Textbook of Medicine, 21st edition, W.B. Saunders, Inc.

Goodkin K, Feaster DJ, Asthana D, et al. (1998, May). A bereavement support group intervention is longitudinally associated with salutory effects on the CD4 cell count and number of physician visits. Clin Diagn Lab Immunol: 5(3); 382-91.

Guyton AC & Hall JE (1996). Textbook of Medical Physiology. Saunders; New York

Harbige LS (1996). Nutrition and immunity with emphasis on infection and autoimmune disease. Nutrition and Health: 10; 285-312.

Hegde HR, Woodman RC, Sankaran K (1999, March). Nutrients as modulators of anergy in acquired immune deficiency syndrome. J Assoc Physicians India; 47(3): 318-25

Herbert TB & Cohen S (1993). Stress and immunity in humans: A meta-analytic review. Psychosomatic Medicine; 55;364-379.

House et al. (1988). Social relationships and health. Science ;241:540-545.

Junker AK, Ochs HD, Clark EA et al. (1986, Sep). Transient immune deficiency in patients with acute Epstein-Barr virus (EBV) infection. Clin Immunol Immunopathol 40(3); 436-446.

Kennedy S, Kiecolt-Glaser JK, Glaser R (1988 Mar). Immunological consequences of acute and chronic stressors: mediating role of interpersonal relationships. Br J Med Psychol; 61(Pt 1):77-85.

Keusch GT & Thea DM (1993). Malnutrition in AIDS. Med Clin North America: 77(4); 795-813.

Kiecolt-Glaser JK, Ricker D, George J (1984). Urinary cortisol levels, cellular immuno-competency, and loneliness in psychiatric inpatients. Psychosomatic Medicine; 46(1): 15-23.

Kiecolt-Glaser JK, Dura JR, Speicher CE et al. (1991). Spousal caregivers of dementia victims: Longitudinal changes in immunity and health. Psychosomatic Medicine; 53;345-362.

Kiecolt-Glaser JK, Glaser R (1992). Acute, psychological stressors and short-term immunological changes. Psychosomatic Medicine; 54;680-685.

Kotze M (1998). Ability of the total lymphocyte count to accurately predict the CD4+ T-cell count in a group of HIV1-infected South African patients. Int Conf AIDS - 1998; 12: 810 (abstract no. 42187)

Laudenslager M, Ryan SM, Drugan RC, et al. (1983). Coping and immunosuppression: Inescapable but not escapable shock suppresses lymphocyte proliferation. Science, 221;568-570.

Learmont J, Tindall B, Evans L, et al (1992). Long-term symptomless HIV-1 infection in recipients of blood products from a single donor. Lancet ;340:863-867.

Leserman J, Jackson ED, Petitto JM, et al. (1999) Progression to AIDS: the effects of stress, depressive symptoms, and social support. Psychosomatic Medicine; 61; 397-406.

Lewi DS, Kater CE, Moreira AC (1995 Mar-Apr). Stimulus of the hypophyseal-adrenocortical axis with corticotropin releasing hormone (CRH) in acquired immunodeficiency syndrome. Evidence for activation of the immune-neuroendocrine system (article in Portuguese). Rev Assoc Med Bras;41(2):109-18.

Lortholary O, Christeff N, Casassus P, Thobie N, Veyssier P, Trogoff B, Torri O, Brauner M, Nunez EA, Guillevin L (1996 Feb). Hypothalamo-pituitary-adrenal function in human immunodeficiency virus-infected men. J Clin Endocrinol Metab ;81(2):791-6

Madhok R, Gracie A, Lowe GD, Burnett A, Froebel K, Follett E, Forbes CD (1986, Oct 18). Impaired cell mediated immunity in haemophilia in the absence of infection with human immunodeficiency virus. Br Med J (Clin Res Ed);293(6553):978-80

McChesney MB & Oldstone A (1987). Viruses perturb lymphocyte functions. Ann Rev Immunol, Volume 5: 279-304.

McDonough RJ, Madden JJ, Falek A, et al. (1980). Alteration of T and null lymphocyte frequencies in the peripheral blood of human opiate addicts: In Vivo evidence for opiate receptor sites on T lymphocytes. J Immunol: 125(6); 2539-43.

Membreno L, Irony I, Dere W, Klein R, Biglieri EG, Cobb E (1987 Sep). Adrenocortical function in acquired immunodeficiency syndrome. J Clin Endocrinol Metab;65(3):482-7.

Mientjes GH, Miedema F, van Ameijden EJ, Hoek AA, et al. (1991). Frequent injecting impairs lymphocyte reactivity in HIV-positive and HIV-negative drug users. AIDS: 5; 35-41.

Momose JJ, Kjellberg RN, Kliman B (1971). High incidence of cortical atrophy of the cerebral and cerebellar hemispheres in Cushing's disease. Radiology 99; 341-348.

Nishijima MK, Takezawa J, Hosotsubo KK et al. (1986). Serial changes in cellular immunity of septic patients with multiple organ-system failure. Critical Care Medicine, Volume 14(2); 87-91.

Norbiato G, Bevilacqua M, Vago T, Clerici M (1996, July). Glucocorticoids and interferon-alpha in the acquired immunodeficiency syndrome. J Clin Endocrinol Metab;81(7):2601-6

Norbiato G, Bevilacqua M, Vago T, Taddei A, Clerici (1997, Oct). Glucocorticoids and the immune function in the human immunodeficiency virus infection: a study in hypercortisolemic and cortisol-resistant patients. J Clin Endocrinol Metab; 82(10): 3260-3.

O'Mahoney JB, Palder SB, Wood JJ, et al. (1984). Depression of cellular immunity after multiple trauma in the absence of sepsis. J Trauma: 24(10); 869-75.

Ornish D (1997). Love and Survival: the Scientific Basis for the Healing Power of Intimacy; Harper Collins; New York.

Pariante CM, Carpiniello B, Orru MG, Sitzia R, Piras A, Farci AM, Del Giacco GS, Piludu G, Miller AH (1997). Chronic caregiving stress alters peripheral blood immune parameters: the role of age and severity of stress. Psychother Psychosom;66(4):199-207.

Polk HC, George CD, Cost K, et al. (1986). A systematic study of host defense processes in badly injured patients. Ann Surg; 204; 282-299.

Sapolsky RM, Uno H, Rebert CS, Finch CE (1990 Sep). Hippocampal damage associated with prolonged glucocorticoid exposure in primates. J Neurosci ; 10(9):2897-902.

Sapolsky RM (1996, August 9). Why stress is bad for your brain. Science 273; 749-750.

Shallenberger F (1998). Selective compartmental dominance: an explanation for a non-infectious, multifactorial etiology for AIDS. Medical Hypotheses: 50; 67-80.

Sridama V, Pacini F, Yang S, et al. (1982). Decreased levels of helper cells: A possible cause of immunodeficiency in pregnancy. New Eng J Med: 307(6); 352-356.

Starkman MN, Gebarski SS, Berent S et al. (1992). Hippocampal formation volume, memory dysfunction, and cortisol levels in patients with Cushing's syndrome. Biological Psychiatry; 32: 756-765.

Stefanski V, Engler H (1998 Jul). Effects of acute and chronic social stress on blood cellular immunity in rats. Physiol Behav;64(5):733-41

Verde TJ, Thomas SG, Moore RW, et al. (1992). Immune responses and increased training of the elite athlete. J Appl Physiol; 73(4); 1494-9.

Verges B, Chavanet P, Desgres J, Vaillant G, Waldner A, Brun JM, Putelat R (1989 Nov). Adrenal function in HIV infected patients. Acta Endocrinol (Copenh);121(5):633-7.

Walton C (1999). What makes a survivor? Continuum 5(5); 16-18.

Williams RC, Koster FT, Kilpatrick KA (1983, November). Alterations in lymphocyte cell surface markers in various human infections. Am J Med: Volume 75; 807-816.

Matthew Irwin
2226 William and Mary Drive
Alexandria, VA 22308


tel/fax:(416) 406-HEAL


Search our web site (400+ pages)
  Site search Web search   search tips