Amir Attaran², Donald R. Roberts¹, Chris F. Curtis³ & Wenceslaus L. Kilama⁴
 

1. Department of Preventive Medicine and Biometrics Uniformed Services University of the Health Sciences Bethesda, Maryland 20814, USA
e-mail: droberts@usuhs.mil
2. Center for International Development Kennedy School of Government Harvard University Cambridge Massachussetts 02138, USA
e-mail: amir_attaran@harvard.edu
3. London School of Hygiene & Tropical Medicine London WC1E 7HT, UK
4. Chairman, Malaria Foundation International; also Chairman-Coordinator, African Malaria Vaccine Testing Network C26/27 Tanzania Commission for Science and Technology Building, Ali Hassan Mwinyi Road, P.O. Box 33207 Dar Es Salaam, Tanzania

Malaria kills over one million people, mainly children, in the tropics each year, and DDT remains one of the few affordable, effective tools against the mosquitoes that transmit the disease. Attaran et al. explain that the scientific literature on the need to withdraw DDT is unpersuasive, and the benefits of DDT in saving lives from malaria are well worth the risks.

However, DDT remains one of the few affordable, effective tools against the mosquitoes that transmit malaria, a plague that sickens at least 300 million and kills over one million, mainly children, in economically underdeveloped areas of the tropics each year. Such a toll is scarcely comprehensible. To visualize it, imagine filling seven Boeing 747s with children, and then crashing them, every day.

Until now, developed countries have grudgingly tolerated the use of DDT against malaria in poor tropical countries; at least 23 countries do so¹. However, this may now be ending. Led by the United Nations Environment Programme, more than 110 countries are negotiating a treaty to "reduce and/or eliminate...the emissions and discharges" of 12 persistent organic pollutants, citing their "unreasonable and otherwise unmanageable risks to human health and the environment."² If it becomes law, the treaty will likely end DDT manufacture, or at least make the supply scarce and unaffordable to tropical countries.

This, in our view and that of nearly 400 colleagues who have signed an open letter to the diplomats negotiating the treaty, is simply dangerous³. The scientific literature is unpersuasive of the need to withdraw DDT; on the contrary, it is clear that doing so risks making malaria control ineffective, unaffordable, or both.

Ecological effects
DDT became emblematic of the toxics movement because of its effects on the non-human environment. Ecological studies have demonstrated that bioaccumulated DDT could cause thinning of eggshells and reproductive failure in birds of prey. The fault for this lies in the massive agricultural use of DDT. Dusting a single 100-hectare cotton field, for example, can require more than 1,100 kg of DDT over 4 weeks⁴.

In contrast, DDT spraying for malaria control is less intensive, less frequent and far more contained. The current practice is to spray the interior surfaces only of houses at risk, leaving a residue of DDT at a concentration of 2 g/m ² on the walls, ceiling and eaves, once or twice a year. Half a kilogram can treat a large house and protect all its inhabitants. Doubtless some fraction of this escapes to the outdoors, but even assuming it all did, the environmental effect is just 0.04% of the effect of spraying the cotton field. Guyana's entire high-risk population for malaria can be protected with the DDT that might otherwise be sprayed on 0.4 km² of cotton in a season⁵. Compared with its agriculture uses, public health uses of DDT are too trivial to merit banning with any urgency.

Tonic...or toxin?
Environment aside, health considerations arise, and with them the dilemma that one man's benefit is another man's risk. Environmentalists in rich, developed countries gain nothing from DDT, and thus small risks felt at home loom larger than health benefits for the poor tropics. More than 200 environmental groups, including Greenpeace, Physicians for Social Responsibility and the World Wildlife Fund, actively condemn DDT for being "a current source of significant injury to...humans."⁶ But five decades of experience with DDT shows that it is highly effective and safe when deployed in house spraying⁷.

Reliance on DDT reached its zenith, and malaria, its nadir, with a campaign to eradicate malaria from large parts of the world in the 1950s and 1960s. The early results were impressive: in less than two decades, spraying of houses with DDT reduced Sri Lanka's malaria burden from 2.8 million cases and 7,300 deaths to 17 cases and no deaths. India and South America achieved similarly impressive reductions, and several countries fully eradicated malaria¹. Even in sub-Saharan Africa, where mosquitoes are most difficult to control, DDT spraying resulted in great reductions in malaria⁸.

Unfortunately, many of these successes were short-lived. American funds, which underwrote the eradication campaign, soon lapsed, and overuse of DDT in agriculture bred DDT-resistant mosquitoes. Back in malaria's grip, Sri Lanka returned to a half a million cases by 1969.

But despite 'resistance' in itself, DDT still works to alleviate mortality and morbidity. Resistance tests work by measuring whether mosquitoes survive a normally toxic dose of DDT. The tests wholly overlook two non-toxic actions of DDT: contact-mediated irritancy⁹, which drives mosquitoes off sprayed walls and out of doors before they bite, and volatile repellency^{10, 11}, which deters their entry in the first place. Both actions disrupt human–mosquito contact and disease transmission.

Data from the Pan-American Health Organization show a strong inverse correlation between malaria cases and rates of spraying houses (1959–1992) in South America, even after DDT resistance became widespread in the 1960s Fig. 1. Here, 'cumulative cases' represent the population-adjusted, 'running' total of cases that exceed or fall short of the average annual number of cases from 1959 to 1979 (years in which World Health Organization strategy emphasized house spraying¹²). Cumulative cases increase considerably in later years, coincident with a sharp decrease in rates of spraying houses.

This inverse correlation is readily understandable because it is so biologically plausible. For mosquitoes, DDT is a toxin, irritant and repellant all rolled into one chemical. All three properties decrease the odds of being bitten by mosquitoes, and toxicity particularly reduces the odds that parasite-bearing mosquitoes will survive to infect others. Lowering these odds slows disease propagation by second- or higher-order relationships and therefore is very important^{13, 14}. Indeed, renewing the spraying of houses with DDT, as Ecuador did in the early 1990s, rapidly decreases case rates⁵.

This body of evidence is so indisputable that even environmental groups such as Physicians for Social Responsibility concede that DDT is "highly effective" in malaria control¹⁵. Campaigning for a DDT ban given this benefit would seem politically difficult unless one alleged even greater health risks associated with its use, which is precisely what environmentalists do. Recent bulletins from Physicians for Social Responsibility and the World Wildlife Fund cite animal studies indicating involvement of DDT in neurological and immune deficits, and epidemiological studies linking DDT to human cancers and endocrine-disrupting effects, such as reduced lactation^{15, 16}.

In this kind of 'balance of risks' paradigm, the evidence must be scrupulously weighed. Although the International Agency for Research on Cancer rates DDT as a possible human carcinogen (along with, notably, several pharmaceutical drugs), not one case-control study of DDT's human carcinogenicity has been affirmatively replicated. Breast cancer furnishes the clearest example: the first study to correlate DDT exposure with statistically elevated risk¹⁷ has now failed to be replicated at least 8 times^18-25, and of these later studies, some found exposure to significantly reduce risk^{24, 25}. Much the same can be said of studies indicating involvement of DDT in multiple myeloma, hepatic cancer and non-Hodgkin lymphoma^{26, 27}.

That DDT interferes with maternal lactation is also questionable. The leading study to correlate DDT metabolites in breast milk with unexplained, premature weaning²⁸ does not reach statistical significance unless the data are first 'adjusted' for potential confounders, but the authors did not disclose how their adjustment was done, and other labs have yet to confirm the result²⁸. Lactation's many social, nutritional and cultural influences make inferences difficult, but even if DDT really abridges lactation, the authors found a "lack of any detectable effect on children's health."²⁸

With such weak evidence of harm to human health, one must decide whether to set policy as a precaution and ban DDT based on animal studies. Ordinarily, this makes sense (given the alternative of experimenting on humans with toxins), but not for the spraying of houses with DDT. Acting with precaution because there are risks in animals, and thus denying people the known health benefits of malaria control, is very unethical: house spraying exposes millions of people to DDT, any of whose health can be studied, making extrapolations from animal studies unnecessary. Proper case-control studies should be done before policy is cast in treaty law.

Indeed, if precaution is relevant, it favors spraying houses with DDT, because it is affordable or effective where other interventions may not be. Cost data from India show that, even using DDT alone, the entire national malaria-control budget is sufficient to protect only 65% of high-risk persons. Switching to malathion, the next-cheapest alternative, reduces that coverage to 21%, which leaves 71 million more persons unprotected²⁹. House spraying also has the advantage that it protects whole families, which is sometimes overlooked in comparing it with insecticide-treated bed-nets, which protect only one or two people at a time³⁰. Simply put, there are too few economic studies to determine with certainty whether bed-nets are more or less cost-effective than DDT house spraying³¹. However, the fact that spraying houses with DDT can lower the prevalence of malaria parasitemia in highly endemic African communities to levels below that achieved by bed-nets (less than 5%) indicates that it is careless to treat them interchangeably⁸.

Patience in all things
How then to reconcile DDT's 'Janus-faced' character? Its benefit in alleviating the suffering of malaria, at a reasonable cost, outweighs any reasonable speculation of its health risks. Living with this may not be easy; changing it is harder still.

Above all, rich countries must allow, and even facilitate, poor tropical countries to make choices about DDT freely, and with informed consent. African countries in particular lack the resources to dispatch health experts to the treaty negotiations, and although it provides financial assistance, the United Nations Environment Programme has declined to assist with this, or even to provide a translator when French- and English-speaking diplomats meet to discuss DDT. The resulting lack of knowledge suffocates debate. At worst, threats are used, as Belize learned when the US Agency for International Development demanded that it stop using DDT.

Such arm-twisting is as lamentable as it is effective. Highly indebted poor countries must of necessity rank poverty reduction over environmental orthodoxy, and stimulating growth and foreign investment will require nearly eliminating malaria from economically productive zones. This is essential for development in sub-Saharan Africa, where malaria subtracts more than one percentage point off the gross domestic product growth rate, for a compounded loss (since 1965) now reaching up to $100 billion a year in foregone income³².

Seen in this way, the insistence to do without DDT is 'eco-colonialism' that can impoverish no less than the imperial colonialism of the past did. Sub-Saharan Africa, which never experienced much spraying of houses with DDT, should consider starting this. South Africa, Swaziland and Madagascar, among others, run successful DDT-spraying programs and prove it can be done^{1, 33}.

At present, the United Nations Environment Programme mandate to "reduce and/or eliminate" DDT probably cannot be accomplished safely, without causing extra disease. As 'preachers of precaution', environmental groups and rich country governments should start by committing at least $1 billion annually to roll back malaria in Africa. That is the sum requested by African leaders at their first-ever Malaria Summit earlier this year³⁴. Meeting this request is a small price to pay for respecting the lives of the poor, and will bring us much closer to no longer needing DDT. Denying it, on the other hand, is to again embrace indifference, and the pursuit of environmental goals, on the backs of world's sickest and poorest.