Banned

by Frederick Rowe Davis

  In the absence of treatment, there were no significant differences in the samples taken, which clearly affirmed that the paired sampling technique was valid. Scientists made a consistent effort to reduce variation by rigidly controlling the sampling technique so that differences due to the treatment might be detected. To determine the significance of the differences in the samples, researchers employed the student’s t-test. They used P values to denote levels of significance (with a value of 0.05 or less considered significant). They clarified the effects of individual treatments with these methods and demonstrated residual or cumulative effects due to routine treatment by comparing graphically the populations in the treated and check (control) ponds throughout the period of treatment.71 Of all the wildlife studies of the effects of DDT, the PHS’s larviciding experiments most faithfully represented the most sophisticated ecological methods at the time.72

  Careful methodology produced clear results. Individual treatments in shallow ponds with a sand bottom at rates of 1 to 2 pounds of DDT per acre killed numerous aquatic animals, including invertebrates like dragonflies and mayflies, as well as fish. Even after they reduced the spray concentrations, PHS officials recorded high mortality rates for nontarget aquatic insects of many orders. Although kills were more pronounced at higher dosages, there were kills at all dosages (even slight kills at 0.025 pound per acre after the first few treatments).

  Meticulous observations clarified the effect of DDT on nontarget aquatic insects. By comparing the effects of DDT spray to the effects of fuel oil alone, the PHS scientists justified the high mortality rates: “The over-all results suggest that 0.05 pound and 0.025 pound of DDT per acre in fuel oil kills only a fraction as many surface forms as do applications at 0.1 pound per acre, and that fuel oil in itself kills numerous forms.”73 The PHS officials suggested that treatments at low concentrations might prove less harmful to nontarget insects than traditional methods of mosquito control: “It may be that 0.025 to 0.05 pound of DDT applied in 1 gallon of fuel oil per acre will kill considerably less insect life than the regular routine oiling at 15 to 40 gallons per acre which has been used for mosquito control in the past.”74

  In addition to the gross observations of mortality, scientists took quantitative surface samples in many of the treated and check ponds. Although a few of the samples showed significant changes in the insect populations in particular ponds, none of these changes were consistent when ponds were compared. The greatest changes in the number of insects occurred in the untreated check pond, but researchers attributed this change to sampling error. Nevertheless, scientists were able to map certain population trends using the data from quantitative sampling. Chironomids (midges) suffered the highest rates of mortality, particularly in a pond treated with a solution of fuel oil and 0.05 pound DDT per acre. Yet the total population of surface forms increased in treated areas. Specifically, nematodes, oligochaetes, and copepods increased in the treated ponds at the wildlife refuge, suggesting that DDT reduced their predators.75

  Researchers hesitated to draw larger conclusions, for example, what these changes might mean for other organisms or the ecosystem as a whole. They did suggest, however, that some of the species that suffered serious declines might represent sources of food for fish and that the forms whose population exploded would not serve as a substitute food source. Nevertheless, they qualified this pessimistic suggestion: “Reductions noted to date, however, have not been sufficient to affect the breeding stock, and since treatment is in localized areas, it is probably not sufficient to seriously limit the fish population by restriction of the food supply.”76 Other PHS biologists would take up the question of the effects of DDT on higher forms, but the potential impact of DDT on the food sources for fish remained central to their research.

  Researchers also examined the effects of larviciding with DDT on birds and mammals at the Savannah River National Wildlife Refuge, beginning shortly after the study of aquatic invertebrates and fish and to a considerable degree sharing the same sophisticated methodology as that study. The study of birds and mammal effects concentrated on about 815 acres of the refuge, which were larvicided during the summer of 1946 with DDT (in a 20 percent solution in highly methylated naphthalene) at the rate of 0.1 pound of DDT per acre.

  Before they sprayed, researchers mapped the ten islands in the study and set out numbered stakes in rows 100 feet apart with 100 feet between the stakes in each row to form a grid for recording the census of singing male birds and the live trapping of mammals. In an effort to build upon earlier research, which emphasized the analysis of the effects of single, high dose treatments of DDT, the PHS biologists modified their approach for light treatments on a regular basis over the extended period required for effective larviciding. Then they counted singing males, on a weekly basis, beginning on March 26, 1946, and continuing until August 8, when the breeding season had concluded for most species. Although the number of singing males increased on both sprayed and unsprayed islands, the arrival of new migrants contributed to this change. In both sprayed and unsprayed areas, the number of singing males rose from a low at the beginning of the nesting season, fluctuated slightly during the season, and fell as the end of the season approached. PHS biologists concluded: “The absence of a sudden drop or a gradual decline in the population of the sprayed area indicates that the DDT spraying did not affect the population to any appreciable extent.”77

  Mammals seemed to be similarly unaffected by routine larviciding with DDT. The most common mammal was the cotton rat (Sigmodon hispidus), a large herbivorous rodent, which was unlikely to feed on insects. Given that the rates of recapture were similar (21.1 percent on the unsprayed areas and 25.4 percent on the sprayed areas), researchers concluded that the activity of the rats and their rate of mortality were about the same on both areas, suggesting that DDT had no apparent effect on the rodent population. Moreover, they trapped only half as many cotton rats on the unsprayed area as on the sprayed area, giving rise to a provocative conclusion: “Thus, DDT had no apparent effect on the reproductive potential of the rats on the sprayed area. Judging by trap catches, the potential was slightly greater in rats of the sprayed area.”78 Researchers also studied rabbits, raccoons, and cotton rats through sight observations on daily drives through the sprayed and unsprayed areas. From 174 observations of immature rabbits on the unsprayed areas and 244 observations on the larger sprayed areas, the biologists concluded that there was no significant difference after allowing for the greater size of the sprayed area. This information did not provide rigorous support for their conclusion, however: “DDT, then was not interfering with the reproductive capacity of the rabbits in the sprayed area as indicated by these counts.”79

  Despite the methodological rigor of the PHS studies of the effects of DDT larviciding on wildlife, the studies could not effectively evaluate nonacute or chronic effects of low concentrations of DDT on wildlife. As in so many other early experiments, these studies focused on a single endpoint—the death of the organism—and anything short of death failed to register in the analyses. In the years following World War II, toxicologists were just beginning to develop methods of analysis that would enable them to evaluate the effects of chronic exposures. Such effects included carcinogenicity and reproductive effects as well as subtle neurotoxic effects. In the case of DDT, scientists were unable to discern a pattern in the appearance of these effects, except that at high doses most animals experienced neurological symptoms prior to death. Fitzhugh did report tumorigenic activity in mice and others noted that DDT could be passed from mother to offspring via milk, but neither of these isolated reports created a clear picture of the toxicity of DDT to all organisms. Studies of the toxicity of DDT to humans were no more definitive.

  Some of the earliest studies of the toxicity of DDT were conducted on human subjects. Scientists wished to rule out potential hazards of occupational and casual exposure to DDT before the chemical reached the general public. The NIH and the PHS took particular interest in the human
health effects of DDT. Neal, Oettingen, and others conducted inhalation experiments on two adults. The subjects underwent extensive tests before and after exposure. Tests included: pulse, blood pressure, respiratory rate, size of pupil and appearance of eyegrounds and conjunctivae and presence of nystagmus, inspection of the throat, steadiness tests (finger-nose test and steadiness of extended fingers), and biceps and Achilles reflexes, urine tests (to determine albumin, sugar, urobilin, urobilinogen, cellular constituents, pH, casts, and specific gravity), blood analysis (red blood cells, hemoglobin, white blood cells, and differentials), and psychophysiological tests (including mental alertness).

  From this extensive battery of tests, Neal and others found no subjective systemic manifestations of DDT intoxication. In addition, results of the physical examination were negative, and the scientists recorded no physiological aberrations other than a drop in hemoglobin and red blood cell count. The NIH scientists summarized their results as follows: “The experiment shows, further, that inhalation of 3.1 to 4.0 mg of DDT in the form of DDT aerosol for 1 hour daily, on 6 consecutive days causes no subjective or objective symptoms in human subjects.”80 In a second series of experiments, researchers raised the exposure level to three times the original experiment. The same two subjects received the higher level beginning four weeks after the original experiment. Again, scientists completed an extensive medical examination before and after exposure. As with the first experiment, none of the tests revealed clinical signs of toxicity despite DDT deposits on the vibrissae (hairs) in the nose of one subject and coating the hands and forearms throughout the entire experiment. Thus the scientists concluded that an exposure to a total of 124.8 grams of a DDT aerosol “produces no toxic effects in human subjects and should offer no serious health hazards if used under conditions required for its use as an insecticide.”81

  Several important studies of the effects of DDT on human subjects were conducted in Great Britain. One researcher completed numerous experiments with volunteers wearing undergarments impregnated with 1 percent DDT (dry-weight basis). This research included also a small group of technicians engaged in laboratory work and bulk impregnation, which brought them into contact with DDT. None of the fifty-eight men manifested symptoms suggestive of toxic absorption, although a few had slight, transient attacks of dermatitis, which may or may not have been caused by DDT. From these results, the researcher concluded that soldiers under battle conditions could safely wear garments impregnated with DDT as a deterrent to lice. He also cited, without references, comparable studies of men engaged in spraying DDT conducted by Americans.82

  Another British researcher placed two male subjects in an octagonal chamber (six feet across, six feet high) for two forty-eight-hour periods with an interval of forty-eight hours in between. The walls of the chamber were painted with a distemper (film). For the first period of exposure, the distemper did not contain DDT, but in the second exposure the distemper contained 2 percent DDT. The subjects wore shorts only and were forced to sit so that a large portion of their skin came into continued contact with the oily film on the walls. Medical examinations (conducted every twenty-four hours) included clinical neurological examination, electro-encephalograms, hematological examination, and urine analysis, as well as notes of subjective phenomena.83

  The extensive examinations revealed no measurable changes during the control period, but there were significant changes during the period of exposure to DDT. Subjectively, both subjects felt eye pain, tiredness, heaviness, and aching limbs and became extremely irritable and disinterested in work of any kind. They felt unable to tackle the simplest mental task (although one subject was able to complete mathematical problems with normal precision). Both subjects suffered intense joint pain, and one had to spend a day in bed. In addition to these pronounced subjective findings, there were slight neurological effects: reflexes diminished, auditory acuity altered, one subject had peripheral anesthesia, and the other had a fine tremor.84 This kind of intense exposure stimulated neurological effects that were not recorded in other studies of human subjects.

  Another British study of the effects of DDT in humans was conducted at the Royal Naval School of Tropical Hygiene. Researchers followed the health of fifteen men attached to the school, who were for many months heavily and continuously contaminated with a 5 percent solution of DDT in kerosene. A variety of clinical and special studies (including renal and liver function, blood investigation, general demeanor, and labor output) did not show any ill effects associated with DDT. In another experiment, researchers exposed six human volunteers for 27.5 hours over five days in an experimental room to a continuous-phase aerosol of DDT. And in yet another experimental room they treated a group of five subjects in the same way at night, and/or continuously for three months. None of the volunteers exhibited ill effects. Royal Naval scientists concluded, “DDT when used as an insecticide, with reasonable intelligence and the precautions normal to the use of modern insecticides, is harmless to man and animals.”85

  In the U.S., besides Neal’s research, another scientist for the PHS conducted much of the analysis of the toxicity of DDT to humans. Based in Savannah, Georgia, Wayland J. Hayes, Jr., was chief of the Toxicology Section of the Technology Branch of the Communicable Disease Center of the PHS. In 1949, Hayes and two colleagues analyzed human fat for the presence of DDT. Previous studies had reported that both humans and animals excreted degraded DDT (metabolites) in urine.86 No systematic study had been published save for the research of Laug and colleagues, who used a technique that made it impossible to determine the quantities of DDT (or metabolites) present in the original fat samples.87 Hayes and his colleagues determined that the majority of samples contained a large proportion of degraded DDT (DDE, a metabolite of DDT), but they could not determine whether DDE was present because DDT residues degraded on plant products prior to ingestion, during digestion, or after deposition in human adipose (fat) tissue. Nevertheless, they called for a reconsideration of the possible health hazards associated with the widespread use of DDT.88 Years later, when it became clear that DDT was wreaking havoc to the reproductive systems of birds of prey, DDE emerged as the culprit.

  In late 1953, Hayes initiated a comprehensive analysis of the toxicity of DDT to humans. With prison volunteers as subjects, Hayes designed experiments to study possible clinical effects at different dosages, the relation between oral dosage and storage of DDT and metabolites in adipose tissue, and the relation between oral dosage and urinary excretion.89 For periods of up to eighteen months, each of the fifty-one volunteers (“with full knowledge of the plan of the study and with complete freedom to withdraw at any time”) consumed 0, 3.5, or 35 mg of DDT every day. Hayes set the dosages according to O. Garth Fitzhugh’s estimate of the daily ambient quantity of DDT consumed by an average adult: 1.75 mg. Hayes set the volunteers’ dosages at roughly 20 times and 200 times this amount. At the end of the study, Hayes and his colleagues commented: “During the entire study, no volunteer complained of any symptom or showed, by the tests used, any sign of illness that did not have an easily recognized cause clearly unrelated to exposure to DDT.”90 Hayes’s research on the storage of DDT in humans showed that after one year humans reached a threshold, after which they accumulated no more DDT. He also found that human subjects excreted 20 percent of the DDT administered as DDA, another metabolite of DDT, like DDE, in their urine. Hayes and his colleagues concluded: “The results indicate that a large safety factor is associated with DDT as it now occurs in the general diet.”91 As in the case of so many of the early evaluations of the toxicity of DDT, scientists defined safety as the absence of acute toxicity or clinical effects. Nor did Hayes consider differences between various kinds of exposure. His ingestion study had no bearing on inhalation toxicity or dermal exposure. By 1956, when Hayes’s study was published in the Journal of the American Medical Society, other scientists, doctors, and wildlife biologists alike were strongly criticizing DDT.92 Despite such currents, Hayes continued to defend its use and saf
ety.

  Historian Edmund Russell has noted: “There were bombs before the atomic bomb, but the atomic bomb placed the attack against human enemies on a new plane. There were drugs before penicillin, but penicillin placed the attack against bacterial diseases on a new plane. There were insecticides before DDT, but DDT placed the attack against insect pests on a new plane.”93 Having examined in considerable detail many of the studies of the toxicity of DDT to target insects, laboratory animals, wildlife (including nontarget insects), and humans, we can now add that DDT not only transformed the attack on insects but also significantly influenced how scientists evaluated the toxicity of new chemicals. During and immediately after World War II, scientists scrutinized DDT intensely. It is safe to say that no chemical before had received such extensive study from such a wide range of scientists: economic entomologists, laboratory scientists, wildlife biologists, public health officials, and doctors.

  Several trends link these disparate investigations. Most of the research focused exclusively on acute effects of DDT. Even the few studies that attempted to address chronic effects generally overexposed the test subjects to DDT (the one major exception being the two-year studies on mice, which also used fairly high doses of DDT). Because of the short duration of most experiments, few toxic effects developed—except at high levels of exposure. Researchers concluded, therefore, that DDT was not harmful. Here the major exception seems to be laboratory studies in which numerous clinical and subclinical effects were noticed but dismissed or minimized in various ways.