¿ìè¶ÌÊÓƵ

Does HIV cause AIDS? A mountain of scientific evidence amassed over
the past ten years says emphatically that it does. Yet a small number of
journalists and dissident scientists continue to question the link. Last
month the debate reached new heights, with the science journal Nature condemning
The Sunday Times for misrepresenting the facts and The Sunday Times accusing
Nature of suppressing evidence that conflicts with the established view.

Crucial to the dissidents’ campaign is the history of the AIDS epidemic
in Africa. They would have us believe that this epidemic is a myth. By contrast,
the WHO says Africa currently bears the greatest burden of the disease.
According to the WHO’s mid-1993 estimates, sub-Saharan Africa accounts for
almost two-thirds of all HIV infections, and more than two-thirds of all
AIDS cases.

Western images of Africa have always been oversimplified. The continent
is often discussed in the West as though it were one country, not some fifty
nations. On television Africa is either a game park or an apocalyptic vision
of famine and civil war. AIDS has suffered from the same kind of oversimplification:
it is either ‘wiping out’ the entire continent, or it scarcely exists at
all.

Who is right? Is the scale of the problem being overstated, or is clear
evidence denied? Here, we try to disentangle the facts from the fiction.

MYTH ONE There is no AIDS in Africa. AIDS is nothing more than a new
name for old diseases

A virus that destroys immunity is likely to have a complex epidemiology,
for the people it strikes will succumb to a variety of infections and illnesses.
The illnesses that have come to be associated with AIDS – malnutrition,
diarrhoeal diseases and TB – have long been severe burdens in Africa. So
how can we be sure these illnesses are the result of HIV infection?

The answer, say health workers on the ground, lies in the kind of people
who are now succumbing. High mortality from diseases like TB is no longer
confined to the elderly and malnourished. ‘There are substantially increased
numbers of young and middle-aged adults dying where we know HIV is prevalent,’
says Klinton Nyamuryekung’e, a doctor working in public health with the
African Medical and Research Foundation in Dar es Salaam, Tanzania. ‘This
is not something we have seen before.’

Further evidence that AIDS is new comes from the impact on the economic
infrastructure of the African countries. Peter Doyle, head of South Africa’s
independent AIDS Modelling Group and chief actuary for a life insurance
company, explains that companies issuing group policies to employers have
taken the unprecedented step of withdrawing business from several southern
African countries because mortality rates in young adults there have doubled.

These deaths are linked with HIV, says Doyle; they are not relabelled
deaths from older diseases. ‘Your cancers and malaria are still there, but
there is suddenly this new cause of death called AIDS.’

Even more obvious is the effect of these mortalities on already fragile
economies. ‘In Tanzania our National Bank of Commerce has started a prevention
programme after it lost key personnel to AIDS,’ says Nyamuryekung’e. ‘A
country which has a limited base of skilled personnel cannot afford to
lose such individuals.’

Outside the cities there is a different problem, he says. ‘In the small
towns and villages, we are losing people who are needed for production
of food . . . Others who are not unwell have to spend time tending the
sick. As a result plantations and farms are not as well maintained as they
were some years ago.’

Epidemiological studies tell the same story. In 1983 there was no AIDS
in Abidjan, Cote d’Ivoire, a city of two million in West Africa. But by
1988 hundreds of people with AIDS were turning up at hospitals and over
the five years fatality rates among young men and women in hospital doubled
and trebled respectively.

Simultaneously, the city’s mortality figures mushroomed for young adults
and by 1989 AIDS had become the leading cause of death. Migrant Africans
who settled in Europe also began to develop AIDS. So far, there have been
over 350 reported cases of AIDS in Africans in Britain. It is hard to imagine
that the doctors reporting these cases are mistakenly diagnosing poverty
or malnutrition as AIDS.

MYTH TWO HIV is not pathogenic in Africa

Citing anomalous findings from unpublished studies – reports claiming
that HIV-positive people are less likely to die than those who are HIV-negative,
for example – The Sunday Times concludes that HIV may not be the cause of
AIDS and that being HIV-positive may only be a marker for something else.
Such reports are invariably unsupported by details of how the testing was
done, or whether there was any follow-up monitoring of the people. Nor are
they subjected to the kind of critical peer review normally required for
scientific studies.

Certainly, research into the epidemiology and natural history of HIV
in Africa is difficult. It requires meticulous laboratory support and rigorous
monitoring of infected populations. Even so, research in Uganda sponsored
by Britain’s Medical Research Council reveals an unequivocal link between
HIV and mortality in the country’s rural southwest, where there is an advanced
epidemic.

Researchers found that children and young adults – aged 13 to 44 years
– who were infected with HIV were sixty times more likely to die within
the two-year follow-up period than were uninfected adults. Nearly half the
community’s adult deaths were attributable to AIDS, even though only 5 per
cent of living adults were infected with HIV. The link between HIV and premature
death was as strong as that seen between smoking and lung cancer. Nor could
it be argued that the infected people died because of fear of their infection:
few of them took up the option of knowing their test result.

Findings are similar elsewhere in Africa. One study, of Rwandan women
whose HIV status came to light during pregnancy, found that the infected
women were many more times likely to die in the two years following their
pregnancy than their counterparts who did not test HIV-positive. The same
was true for patients with TB in Nairobi who tested HIV-positive.

All these studies used conventional blood tests for HIV and were subjected
to peer review before publication. Moreover, observations showed that factors
such as age, sex and socioeconomic status barely influenced the link between
HIV infection and death. If being HIV positive is only a marker for something
else then that something has to be both extremely elusive and highly lethal.

MYTH THREE Results of AIDS research in Africa are unreliable because
all HIV testing in Africa is unreliable

In questioning the validity of AIDS research in Africa, some journalists
have seized on stories of people testing positive for HIV and later testing
negative. They have also made much of the fact that some of the earliest
predictions of the likely impact of HIV on Africa were excessively pessimistic.

The main tests for HIV infection are based on detecting antibodies to
the virus in people’s blood. Some of the first tests developed in North
America hit problems in Africa in the mid-1980s. The tests reacted with
other antibodies, including those against malaria parasites. As a result,
infection levels in parts of Africa, notably rural Kenya, appeared to be
much higher than they really were.

These elevated levels of infection were reported by researchers in
the US in 1984. Although the findings were later retracted, the damage had
been done. Wild (and inaccurate) apocalyptic predictions followed, and it
became hard to dispel the notion that malaria gives false positives, even
though the HIV-antibody tests used since 1986 very rarely do so.

In South Africa, Doyle explains how ‘gross’ exaggeration in the early
years has blinded commentators to the severity of the real epidemic that
now confronts the country. ‘Early on there were a lot of people who for
political reasons said the population would be wiped out, or that the population
growth rate would go negative, or that 50 per cent of the population would
have AIDS,’ he says. ‘Now that this is not happening, people are saying
AIDS is a myth.’

Doyle became involved in projections to debunk the wildly exaggerated
claims. Now he finds himself stressing just how serious the real, if lower,
levels of infection are. On the basis of data from anonymous testing of
pregnant women, Doyle’s group estimates that some 300 000 people out of
South Africa’s total population of 35 million are infected with HIV today.
Government estimates are roughly the same. Unless conditions change, the
number will treble to reach about 1 million people who are HIV-positive
by 1996.

Yet prophecies of populations shrinking in Africa are now seen as unlikely
to be fulfilled. The scale of the problem is still much greater than in
Europe or North America but the situation varies enormously from place to
place, even within a single country. As with other sexually transmitted
diseases, the spread varies according to how frequently people change partners
and whether they change partners within a well-defined social group or more
widely. Infection is always more frequent in towns and roadside trading
centres, and it is possible to find tenfold differences in adult infection
rates within a few miles. Disease and death driven by HIV will vary similarly.

Any biological method for diagnosing disease or infection will occasionally
produce ambiguous results (see This Week, 18 September). HIV tests are no
less reliable than any other. A skilled laboratory technician can almost
always distinguish genuine infection from a false result by repeating the
test on a fresh sample or by using a combination of tests. The difficulty
in Africa is that few laboratories have the resources or expertise to do
this routinely.

Consider the case of Tanzania. Like most African countries it has a
national HIV test reference centre which takes part in the WHO’s quality
assurance schemes. There are also four sophisticated consultant hospitals
where HIV testing is accurate. ‘We know this from exchanges of specimens
with WHO reference laboratories in Europe,’ says Nyamuryekung’e. European
laboratories check the results obtained in Tanzania and the findings mostly
agree; the same is true for checks made on European samples by the Tanzanian
laboratories.

However, outside these centres the picture is less certain. Poverty,
the cost of tests and unfavourable conditions for transport, storage and
supervision all conspire to make HIV testing less accurate in the small
hospitals, says Nyamuryekung’e. ‘Certainly errors can be made there and
false positives and negatives do occur. Our priority in the small hospitals
is to ensure safety of the blood supply (as opposed to diagnosing infection
or producing scientific data), and so we make sure the tests we select are
highly sensitive.’

The WHO’s estimates of overall numbers of HIV-1 infections and AIDS
cases are based on extrapolations from the most reliable data on HIV infection
in Africa. The figures are produced by a mathematical model called EPIMODEL,
usually using a period of ten years as an estimate of the interval from
infection to the onset of AIDS. Researchers reject data if the quality of
sampling or testing is uncertain and choose the lower figure if two estimates
give different results.

Some people are puzzled by apparent discrepancies between the figures
for HIV and AIDS in Africa and the figures for other continents. When people
examine the ratios of reported AIDS cases to the estimated number of HIV
infections in North America or Western Europe and compare these to the ratios
in Africa, there seem to be a lot of ‘missing’ AIDS cases in Africa. The
explanation lies with the difficulty of reporting AIDS in Africa. ‘There
is considerable underreporting of AIDS cases in Tanzania for a series of
reasons,’ says Nyamurye-kung’e. ‘Many patients find it impossible to make
the trip to the hospital or they seek other sources of care. The obligation
felt by the doctor to report cases is also weaker, and a reasonable estimate
of the proportion of cases that are reported is between 10 and 20 per cent.’
This may seem a small figure but it is higher than the proportion of other
severe tropical infections, such as cholera and yellow fever, that are officially
reported in Africa.

Nyamuryekung’e and his colleagues estimate from HIV testing of blood
donations and other data that the total number of infected individuals in
the country in 1992 was around 800 000. ‘When we adjust our number of reported
AIDS cases (38 719 to December 1992) to account for underreporting we come
to a figure of around 190 000. That is close to what would be expected given
the level of infection shown by blood testing, the stage of our epidemic
and ratios of AIDS cases to numbers infected in other parts of the world.’

MYTH FOUR AIDS cases are overreported in Africa. Many are not even HIV
infected

Much confusion stems from the different criteria employed for defining
AIDS cases in Africa. Many AIDS cases in Africa have not been confirmed
by antibody tests and some people described as having AIDS test negative
for HIV antibodies. Why?

The answer lies with the history of AIDS surveillance in Africa. When
in 1985 it became clear that there was a major AIDS problem, health officials
needed to establish surveillance so that there could be some estimate of
the extent of the epidemic. As an interim solution in 1985, health officials
established a definition of AIDS that was based just on clinical signs and
symptoms. As there was no need for a laboratory test, even the poorest African
health ministries could ask for reports and track their national epidemic
of AIDS.

Such clinical case definitions are crude tools. They are relatively
insensitive and nonspecific – a person may have symptoms and diseases that
fit the definition without being infected with HIV. But if this causes any
overreporting, it is offset by the much bigger forces leading to underreporting,
say experts like Nyamuryekung’e and Kevin De Cock of the London School of
Hygiene and Tropical Medicine.

Some critics have distorted this problem to argue that HIV cannot be
the cause of AIDS. ‘HIV-negative AIDS’ can look the same as ‘HIV-positive
AIDS’, they say. But such problems are not unique to AIDS. In Africa, misdiagnosis
is common with malaria, TB and other serious diseases that are often diagnosed
without laboratory tests for their causative agents. Yet no one questions
whether Mycobacterium tuberculosis causes TB or plasmodium parasites cause
malaria.

There are many important unanswered questions about HIV in Africa. Information
about the time interval between HIV infec-tion and the onset of AIDS in
African adults is limited. Some believe the interval is shorter than the
one researchers use at present to predict future numbers of AIDS cases.
But what little information exists is open to many interpretations.

Similarly, it is unclear if malnutrition or the coexistence of infections
hasten the progress of the disease. Why is the incidence of HIV so high
in heterosexuals in some African settings, and why does the prevalence of
HIV in populations rise to very different levels in different places? Do
biological differences in the strains of virus in Africa play an important
part in how fast they spread, or in the rate of progression to AIDS?

We are clearly into the ‘long haul’ for AIDS prevention and research
in Africa. On the practical side, HIV prevention must be integrated with
and strengthen (rather than distract from) existing primary health care.
Research into AIDS in Africa should be intensified in the years ahead. Peter
Piot, Director of Research and Development for the WHO’s programme on AIDS,
and others have argued strongly for industrialised countries to invest in
developing Africa’s own research base for this purpose, instead of allowing
teams of expatriate scientists to fly in and conduct ‘safari’ research.
Piot has also gone on record to state that ‘it is truly criminal to tell
our youth that AIDS does not exist, or that it cannot be transmitted from
a woman to a man, and that condoms do not protect against HIV’.

Angus Nicoll is at the Public Health Laboratory Service in London. Further
reading: ‘The contribution of HIV-1 to adult mortality in a Ugandan population’,
AIDS, 1994. ‘Trends in mortality due to AIDS in Africa’, AIDS, p 398, 1991.
‘AIDS case definitions in Africa’, British Medical Journal, p 1185, 1991.

* * *

No end in sight in war of words

For more than two years, The Sunday Times has campaigned to convince
its readers of three claims that run counter to the consensus of medical
and epidemiological research: that HIV may be irrelevant to AIDS; that AIDS
in Africa scarcely exists; and that heterosexuals are not at risk from the
disease. Much of the original stimulus for these claims comes from the views
of Peter Duesberg, a California-based biologist who has long believed that
HIV is not the cause of AIDS (‘AIDS and the innocent virus’, ¿ìè¶ÌÊÓƵ,
28 April 1988).

Certain other British newspapers, including the Sun and the Daily Express,
have toyed with the arguments too, but within the print media only The Sunday
Times has espoused the cause with missionary zeal, publishing more than
twenty articles on the subject since December 1992.

Gaping holes in the logic behind these claims have repeatedly been exposed
by a wide range of scientists and from time to time have been reported in
¿ìè¶ÌÊÓƵ. But the newspaper’s stance has nevertheless triggered parliamentary
questions, the wrath of the WHO and mild bemusement among the scientific
community in the US.

The campaign has now become entirely self-fuelling. Each article contains
some claim which provokes fury from scientists and public health professionals,
who criticise the accuracy and choice of evidence. Each time the newspaper
belittles the role of HIV in AIDS or denies the risk to heterosexuals, lives
are put at risk, they say.

Yet the professionals’ anger is now the meat and drink of the campaign.
The Sunday Times, said to be Britain’s most popular broadsheet newspaper,
is now presenting itself as some kind of pious David against Goliath: a
lone seeker after truth, persecuted by a giant but defensive ‘AIDS industry’.

Now the campaign has taken a new twist. At the end of November, the
newspaper published a long and self-laudatory article by its scientific
editor, Neville Hodgkinson, claiming to ‘set straight the half-truths and
fictions surrounding the disease’. The journal Nature chose to respond with
a two-page editorial criticising the newspaper’s stance and announcing plans
to report The Sunday Times’s coverage of AIDS in its own news section each
week, ‘if only to save readers the trouble of buying it’.

Next came round two. The Sunday Times hit back, more pious than ever.
‘Who is afraid of the truth?’ it asked in an article on 12 December. Steaming
with righteous indignation, it attacked Nature for refusing to publish articles
and a letter by the minority who question the role of HIV in AIDS. ‘Neither
the quality press nor mainstream science would be so sure of themselves
if one magazine had been doing its job properly . . .’ claimed Hodgkinson.
Yet the newspaper has itself repeatedly refused to publish letters and evidence
from scientists who argue against its own line. Hodgkinson says ‘dozens’
of critical letters have been published, but many dozens more have not.
John Maddox, the editor of Nature, defends his decision not to publish the
dissenters’ letter and other material on the grounds that the journal’s
readers are already quite familiar with Duesberg’s views and that ’round
robin’ letters are always avoided for being ‘hortatory rather than reasonable’.

Nature’s running commentary on The Sunday Times’s coverage of AIDS,
which began last month, has alleged that in recent weeks the newspaper
has:

continued to ignore the views of clinical researchers working with AIDS
patients

continued to peddle the views of a ‘small but vociferous group of self-styled
dissidents who do not believe HIV causes AIDS’

recycled an old – and strongly refuted – claim that some HIV-positive
haemophiliacs have recovered after being treated for a blood disorder

reported the discovery of HIV-negative cases of Kaposi’s sarcoma (a
skin cancer often seen in AIDS patients) as if it is a new finding, when
in fact such cases have long been in the medical literature

Last month The Sunday Times defended its coverage of what it sees as
‘anti-HIV’ findings, saying: ‘We have never argued that any of these findings
rule out a role for HIV in AIDS.’

* * *

Myths about disease

People in all cultures subscribe to ‘lay-beliefs’ or myths about health
which may be factually wrong but are highly influential in determining how
people act. Education does not necessarily remove these ideas. In Europe,
for example, many otherwise sophisticated people spend large amounts of
money purchasing patent medicines of no demonstrable benefit.

Studies in East Africa suggest that people have equally false perceptions
about HIV. For example: that the virus is transmitted by mosquitoes (it
is not); that healthy-looking people cannot be infected with HIV; and that
condoms are deliberately impregnated with HIV. Many people attributed their
beliefs to what they had read in the local newspapers.

False beliefs about AIDS and sexual behaviour are not confined to East
Africa. In Europe women may also believe apo-cryphal stories that condoms
can disappear inside them, a belief which is also widespread in Africa.
And the idea that manufacturers deliberately impregnate condoms with HIV
echoes a similar belief expressed in Britain forty years ago, by people
who believed that syphilis was being spread in the same way.

Anthropologists suggest that some of these myths provide people with
reassurance, justifying their continuing with risky sexual practices, such
as having casual partners and avoiding condoms. Could it be that the myths
enjoying currency in some of the British media are an example of the same
kind of denial?

Phyllida Brown