Correspondence on the paper "Agricultural production and malaria resurgence in Central America and India"
Nature Vol 294 26 November 1981 pages 302,388
SIR --- I have read the paper by Chapin and Wasserstrom (Nature 17
September, p.181) with interest. I am disappointed with the
presentation and discussion of the important subject of malaria
resurgence and its relationship to agricultural production.
The authors give a garbled account of the very concept of malaria
eradication and especially of the causes of the relative failure of
this great endeavour. They imply, that the main obstacle to the early
achievement of the planned goal was resistance of Anopheles to
insecticides. This is not so, even though the latter phenomenon played
a significant part in technical problems that the World Health
Organization (WHO) was facing during the late 1960s. The most
comprehensive analysis of the multiple causes of the disappointing
progress of malaria eradication in some parts of the tropical world
was presented by WHO in 1969. It pointed out that administrative,
social, economic and financial factors were largely responsible for
the resurgence of malaria at the end of that decade, especially in
There is nothing "ironical" in the fact that agriculture in India and
elsewhere expanded in regions where malaria incidence decreased
spectacularly. It is precisely because of the improved health
conditions, due to the use of residual insecticides against Anopheles
vectors, that efficient agriculture became possible not only by rich
landlords but also by small farmers. The astounding graph showing
the positive association between the use of DDT in India and the
striking increase of malaria in 1969-77 cannot be accepted in support
of the authors' thesis of a causal relationship between the two
An epidemiological analysis of the resurgence of malaria in India,
such as the one carried out by Akhtar and Learmonth, showed that the
increase in the incidence of malaria in India from 350,000 cases in
1969 to nearly 2 million in 1973 and then to 5 million in 1975 was due
to other conditions with an adverse affect on the standard of
anti-malaria operations in India and in other countries of the
subcontinent. The military conflict with Pakistan, the sharp fall in
the flow of American aid and the temporary food shortages due to bad
harvests took place during that period. There were delays in the
allocation of foreign exchange for insecticides and drugs and there
was some loss of urgency of the malaria control campaign because of
its good results and the emphasis on family planning.
The successful 1965-69 phase of malaria eradication in India left
uncompleted four large areas from which the consequent resurgence
spread. These were the Rann of Kutch in western India, Madhya Pradesh
hill forests, Orissa hill forest tracts and the forested areas of
Assam. Akhtar and Learmonth indicated that densely populated areas,
with extensive irrigation and high agricultural production, have shown
less malaria than other areas during the period 1970-75. On the other
hand, in spite of the continuous use of DDT for agricultural needs,
the amount of malaria in India decreased sharply between 1977 and
1980, thanks to better implementation of control programmes.
A series of valuable studies by Indian experts[3,4] and an Audit
Report of the Agency for International Development attach much less
importance to the problem of insecticide resistance and could be
quoted as an argument against the thesis so eagerly adopted by Chapin
The adverse effects of excessive use of residual insecticides in
agriculture have been known and stressed by WHO for well over a
decade. It is quite untrue that WHO did not urge all countries in the
developing world to decrease as much as possible (without endangering
their food and health programmes) the use of residual insecticides and
to introduce alternative methods of pest control. The emphasis on this
policy was clearly stated in the 16th Report of the WHO Expert
Committee on Malaria and was repeated and emphasized in the 22nd
Report of the WHO Expert Committee on Insecticides and in every
other relevant WHO report.
The serious imputation that the Food and Agricultural Organization
(FAO) policy is being influenced by large insecticide-producing
companies will certainly be answered by those accused of
irresponsibility or other more sinister intentions. The authors use
with evident relish the impressive term "integrated control" although
it is doubtful that they fully understand its practical implications.
Integrated pest management means a combination of chemical, biological
and environmental methods. The complexity of implementation of these
methods is often difficult to fathom by non-specialists, who invoke
the term with more heat than light and over-estimate its universal
feasibility. Integrated control methods are certainly successful in
some areas with important crops but they must be tailored to local
conditions. Their use requires not only a constant assessment of the
size of the pest population but also carries with it the uncertainty
of results should the crops be attacked by more than one
pest. Chemical pesticides are still the main component of most methods
of integrated control and several progressive companies have now
developed insecticides (chlorvinphos) that are particularly suitable
for this purpose.
Integrated pest management is now accepted and promoted by WHO and FAO
wherever scientific agriculture and progressive health programmes
exist. No doubt wider use of this approach is desirable but this
depends largely on the availability of well trained specialists and on
the understanding by farmers of the need for restraint in the use of
chemicals. In this respect the paper by Chapin and Wasserstrom is of
some value, although its sensational presentation (with sub-titles
such as "Deadly link") undermines the credibility of the authors
Member of WHO Expert Panel on Malaria,
1. WHO Official Records of the WHO 176, 106-124 (1969).
2. Akhtar, R. & Learmonth, A. Malaria (Research Paper
No.3, Science Faculty, Open University, 1979).
3. Ray, A.P. J. Communicable Dis. 9, 145-171 (1977).
4. Rao, T.R. Sisir Kumar Mitra meml Led. (Indian National
Science Academy, New Delhi, 1978).
5. Agency for International Development Audit Report
6. WHO tech. Rep. Ser. No.549 (1974).
7. WHO tech. Rep. Ser. No.585 (1976).
SIR --- It is generally agreed among malariologists that agricultural
insecticides have made a contribution to selection for insecticide
resistance in mosquitoes and that such resistance has made a
contribution to the resurgence of malaria in Central America and South
Asia. It is also generally agreed that one should be very careful
before inferring a causal relationship from the discovery of a
correlation between two sets of measurements.
No such care was exercised by Chapin and Wasserstrom (Nature 17
September, p.181-185). They infer that in El Salvador "each kilo of
insecticide added to the environment will generate 105 new cases of
malaria". Taken literally and from their own data this would imply 168
million cases of the disease in a country with a population of 4.3
Chapin and Wasserstrom present three figures with apparently
calculated points and lines but no actual data points. Each graph has
a correlation coefficient marked on it of 0.96 or 0.99, but,
especially in the case of the curvilinear relationships, the reader
has no means of assessing how these coefficients were
calculated. Their Fig. 1 is entitled "Effect of DDT on rice production
in India, 1970-77", as if the correlation between these two parameters
were a simple causal one. In fact, however, it is well known that
other independent factors, notably the introduction of high yielding
varieties, have both boosted rice yield and allowed and/or required
more insecticide usage.
The basis of Chapin and Wasserstrom's curves (Figs 2 and 3) purporting
to relate malaria incidence in India during 1969-77 to DDT usage and
rice production, is unclear. In the figure (see below) I have plotted
the data on malaria incidence issued by the Indian National Malaria
Eradication Programme (NMEP) against the data shown by Chapin and
Wasserstrom (Fig.4) for DDT usage. Certainly both quantities tended
to rise over those years but so did other probably relevant factors
such as irrigation. My graph shows a much less startling relationship
than Chapin and Wasserstrom's because, according to the NMEP data, the
minimum number of cases was higher, the maximum was lower and the
malaria resurgence peaked in 1976 (and is reported to have continued
to fall in subsequent years). No doubt the NMEP figures greatly
under-report the true incidence of the disease, but there seems no
reason to suppose that this under-reporting was greater in the later
years than in the earlier: Chapin and Wasserstrom give no information
about any "correction factors" which they may have applied to the
available data. I conclude that Chapin and Wasserstrom's graphs give a
grossly misleading impression that there is a simple causal
relationship between agricultural insecticides and malaria.
That the relationship is actually more complex is indicated by the
(1) Spraying of cotton crops has the, at least short term, beneficial
side effect of suppressing mosquito populations.
(2) The large tonnage of insecticides used in anti-malaria spraying
has certainly contributed to the selection for insecticide resistance
in mosquitoes, as shown by the fact that withdrawal of this spraying
has been found to lead to a levelling out or decline in the frequency
of resistance genes.
(3) In the Gezira area of intensive agriculture in Sudan, resistance
in Anopheles arabiensis is to malathion which is used in anti-malaria
spraying and not to the other organophosphates used for spraying the
(4) Sri Lanka has very wisely banned the use of DDT and malathion in
agriculture and reserved them for the anti-malaria campaign but has
still had a hard struggle to contain and reverse its resurgent malaria
London School of Hygiene and
SIR -- We are sorry that Professor Bruce-Chwatt disagrees so radically
with our interpretation of the facts surrounding malaria resurgence in
India. In our own defence, however, we would suggest that the military
conflict with Pakistan, the sharp fall in the flow of American aid,
the temporary food shortages that took place there between 1973 and
1975, etc., do not explain the abundant entomological reports of
Anopheles resistance which we cited in our article --- even from such
prosperous agricultural regions as Maharashtra and Gujarat. These
reports, together with official accounts of the deliberations within
WHO and FAO, constitute the major source of evidence upon which we
have based our analysis.
Finally, like Professor Bruce-Chwatt, we appreciate the difficulties
of implementing effective systems of integrated pest management in
tropical areas. It was during the successful development of one such
system in southern Mexico that the idea of writing this article first
occurred to us.
In reply to Dr Curtis, it is unfortunate that in the process of
reproducing our illustrations, many of the data points have become
difficult to discern. Disregarding for the moment the 1979 figure on
malaria incidence in India, however, Dr Curtis's information suggests
an even higher correlation between DDT usage and the spread of disease
than we have calculated. As for the question of decreased
transmission, it may well be true that malaria in India peaked in
1976, but the 1977 figure he cites has been questioned by numerous
specialists and in any case does not contradict our argument about
As for the four points raised in his letter, we offer the following
(1) It is precisely the short-term beneficial effect of insecticides
that encourages cotton growers and public health officials to apply
them. What we have argued, however, is that the disadvantages of
insect resistance soon come to outweigh these rather ephemeral
(2) It is difficult to separate the effects of insecticides used for
malaria control and those used in agriculture. What is clear, however,
is that the decline in resistance after anti-malaria programmes are
discontinued is a limited and unfortunately rare phenomenon.
(3) Although it may be true that Anopheles mosquitoes in Sudan are not
resistant to the organophosphates used on cotton, most countries have
not been so lucky. How does Dr Curtis explain the almost complete and
apparently irreversible resistance among malaria vectors in India,
South-East Asia and Central America?
(4) As the case of Sri Lanka indicates, restricting the use of a
particular chemical does not guarantee that mosquitoes will remain
susceptible to it: application of related (and even unrelated)
compounds is often sufficient to stimulate resistance. Moreover, if
initial success in combating malaria leads to the reduction of
screening and treatment procedures (as commonly occurs), epidemic
resurgence will indeed be exceedingly difficult to control.
New York, New York, USA
As I was saying about the simple correlations.
There is much more still not being discussed, (as far as the correlations supposedly related to larger provinces in India, and as far as net lives vs. DDT as a vector control utility and as an insecticide factor for more net crop growth during the green revo. and all)
Tim, you jumped up about 10 feet higher in my book.