The Lethal Dose

							                    The Lethal Dose (LD) Test - (LD-50 test)
The lethal dose (or LD) test measures the amount of a toxic substance that will, in a single
dose, kill a certain percentage of animals in a test group. "To avoid interference with
results," no painkillers are administered. Each year, about five million dogs, rabbits, rats,
monkeys, and other animals die in lethal dose tests performed in the United States.(1)

Cosmetics and household product tests on animals are not required by law. The Food and
Drug Administration (FDA) requires only that each ingredient in a cosmetic product be
"adequately substantiated for safety" prior to marketing, or the product must carry a
warning on the label that its safety has not been determined. The FDA does not have the
authority to require any particular product test, so contrary to what companies that use
lethal dose tests say, the FDA does not require the test.

During a lethal dose test, the experimental substance is forced into the animals' throats or is
pumped into their stomachs by a tube, sometimes causing death by stomach rupture or
from the sheer bulk of the chemical dosage. Substances also are injected under the skin,
into a vein, or into the peritoneal lining of the abdomen; they are also often applied to the
eyes, rectum, or vagina, or forcibly inhaled through a gas mask.

The lethal dose test does not accurately measure human health hazards and is very crude
and imprecise. Lethal dose test results can be affected by the age and sex of the animals
tested, their housing and nutritional conditions, temperature, time of day and year, and the
exact method used to administer the substance.(2) Different species react differently to
substances, and reactions between individuals of the same species can also vary greatly. For
example, nicotine is lethal to humans at 0.9mg/kg, but lethal dose values of nicotine in dogs
are a staggering 9.2mg/kg; in pigeons, 75mg/kg; and in rats, 53mg/kg.

According to Robert Osterberg of the FDA, "Bureaucrats need an atom bomb to move
them." In response to public pressure, some laboratories have begun reducing the number
of animals used in revised lethal dose or "limit" tests, but the tests remain cruel and
inaccurate. As with the notorious eye irritancy tests, it will take pressure from concerned
consumers and animal advocates to move cosmetics companies into the world of
progressive science. The lethal dose test will meet its overdue demise when enough
consumers, legislators, and ethical scientists demand an end to its use.
Method of Miller and Tainter

Method of Miller and Tainter is very popular. In this method, percentage mortality is
converted into probability units (probits) by referring to appropriate tables and the obtained
probit values are plotted against log-doses to get a log-dose probit graph. Probit 5
(equivalent to 50% mortality) on the Y-axis is interpolated to the X-axis to get log-LD50, the
antilogarithm of which gives LD50

Method of Reed Muench

The methods that have been most widely used in the United States are the Probit method
of Gaddum 2 and Blisss and the so-called Reed-Muench method.6 (It is of interest to note
that the latter method, although independently developed, is essentially the same as the
Behrens7 method which had been in use in Europe prior to its publication by Reed and
Muench.) The Probit method has proved its value in bioassay and it is generally considered
the best to use in exact work, but it has the disadvantage that the computations are
complicated and not too easily comprehended by many workers in problems of bioassay.
The simplicity of the Reed and Muench method has contributed to its popularity. It,
however, does not provide an exact statement of error and lacks other information that is
desirable.

The Wilson-Worcester method

The Wilson-Worcester method, rather recently proposed, uses the logistic in place of the
integrated normal curve for obtaining the probability of an observed response in a sample.
Explicit and not very complicated formulae are given for the determination of the ED 50
point and its standard error and also for determining whether the ratio of two assays at the
ED 50 point will hold, within the sampling error, throughoutthe range of dosage.

Other

Other methods or modification of the Probit method have been proposed for bioassay
problems by a number of workers, among whom are Karber Litchfield and Fertig Berkson,
Wilson and Worcester, Knudson and Curtis and Thompson.