Content uploaded by Martin E Hugh-Jones
Author content
All content in this area was uploaded by Martin E Hugh-Jones on May 07, 2014
Content may be subject to copyright.
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
The Sverdlovsk Anthrax
Outbreak of 1979
Matthew Meselson,* Jeanne Guillemin, Martin Hugh-Jones,
Alexander Langmuir,† Ilona Popova, Alexis Shelokov,
Olga Yampolskaya
In April and May 1979, an unusual anthrax epidemic occurred in Sverdlovsk, Union of Soviet Socialist
Republics. Soviet officials attributed it to consumption of contaminated meat. U.S. agencies
attributed it to inhalation of spores accidentally released at a military microbiology facility in the city.
Epidemiological data show that most victims worked or lived in a narrow zone extending from the
military facility to the southern city limit. Farther south, livestock died of anthrax along the zone’s
extended axis. The zone paralleled the northerly wind that prevailed shortly before the outbreak. It is
concluded that the escape of an aerosol of anthrax pathogen at the military facility caused the
outbreak.
Anthrax is an acute disease that
primarily affects domesticated and
wild herbivores and is caused by
the spore-forming bacterium
Bacillus anthracis. Human anthrax
results from cutaneous infection
or, more rarely, from ingestion or
inhalation of the pathogen from
contaminated animal products (1).
Anthrax has also caused concern
as a possible agent of biological
warfare (2).
Early in 1980, reports appeared
in the Western press of an anthrax
epidemic in Sverdlovsk, a city of
1.2 million people 1400 km east of
Moscow (3, 4). Later that year,
articles in Soviet medical,
veterinary, and legal journals
reported an anthrax outbreak
among livestock south of the city
in the spring of 1979 and stated
that people developed gastro-
intestinal anthrax after eating
contaminated meat and cutaneous
anthrax after contact with diseased
animals (5-7). The epidemic has
occasioned intense international
debate and speculation as to
whether it was natural or
accidental and, if accidental,
whether it resulted from activities
prohibited by the Biological
Weapons Convention of 1972 (8).
In 1986, one of the present
authors (M.M.) renewed
previously unsuccessful requests to
Soviet officials to bring
independent scientists to
Sverdlovsk to investigate. This
resulted in an invitation to come to
Moscow for discussions with four
physicians who had gone to
Sverdlovsk to deal with the
outbreak (including another of the
present authors, O.Y., who was a
clinician in the intensive care unit
set aside to treat the victims). In
1988, two of these Soviet
physicians visited the United
States, where they gave formal
presentations and participated in
discussions with private and
government specialists. According
to their account, contaminated
animals and meat from an
epizootic south of the city starting
in late March 1979 caused 96 cases
of human anthrax with onsets from
4 April to 18 May. Of these cases,
79 were said to be gastrointestinal
and 17 cutaneous, with 64 deaths
among the former and none among
the latter (9).
The impression left on those of
the present authors who took part
in the U.S. meetings (J.G., A.L.,
M.M., and A.S.) was that a
plausible case had been made but
that additional epidemiological and
pathoanatomical evidence was
needed. Further requests by M.M.
for an invitation led to an on-site
study in Sverdlovsk, initiated there
in June 1992, and a return visit in
August 1993.
Starting in 1990, several
articles about the epidemic
appeared in the Russian press (10).
These included interviews with
Sverdlovsk physicians who
questioned the food-borne
explanation of the epidemic and
with officials at the military
microbiology facility. These
officials said that in 1979 they had
been developing an improved
vaccine against anthrax but knew
of no escape of anthrax pathogen.
Late in 1991, Russian President
Boris Yeltsin, who in 1979 was the
chief Communist Party official of
the Sverdlovsk region, directed his
Counsellor for Ecology and Health
to determine the origin of the
epidemic (11). In May 1992,
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
Fig. 1. Time course of the epidemic:
onsets of fatal cases by week. The
first week begins on 4 April 1979, the
date of the first onset we recorded.
Lighter shading represents cases for
which the onset date is unknown and
is estimated by subtracting 3 days
from the date of death.
Yeltsin was quoted as saying that “the
KGB admitted that our military
developments were the cause” (12).
No further information was provided.
Subsequently, the chairman of the
committee created by Yeltsin to
oversee biological and chemical dis-
armament expressed doubt that the
infection originated at the military
facility and stated that his committee
would conduct its own investigation
(13). The results of that investigation
have not yet appeared.
Pathoanatomical evidence that the
fatal cases were inhalatory, recently
published by Russian pathologists who
performed autopsies during the
epidemic (14-16), is summarized in an
earlier report from the present study
(17). Here we report epidemiological
findings that confirm that the pathogen
was airborne, and we identify the
location and date of its escape into the
atmosphere.
Table 1. Case data. Case numbers for fatalities are as they appear on the
administrative list. Case numbers for survivors are arbitrary. Days of onset
and death are counted from 1 April 1979. Abbreviations: O, onset; D, death;
R, residence; W, workplace; *, unidentified man; ?, not known; ma, mid-
April; s, survivor; c, cutaneous survivor; +, in high-risk zone; –, outside high-
risk zone; a, had two residences, one in Compound 32; p, pensioner; r,
daytime military reservist at Compound 32; u, unemployed. Patients 25, 29,
48, and 87 were home on vacation during the first week of April.
—————————————————————————————————
Case Age/sex O/D R/W Case Age/sex O/D R/W
no. no.
—————————————————————————————————————
* ?/m ?/? ?/? 51 31/m 10/15 –/?
32 40/m ?/? –/? 40 37/m 12/15 +/+
67 26/m ?/? +/+ 36 68/f ?/16 a/p
68 32/f ?/? +/+ 35 52/m 13/16 +/+
8 60/f ?/8 +/+ 34 43/m 14/16 –/+
18 38/m 6/8 –/? 38 69/f 14/16 +/p
16 40/m ?/9 +/+ 39 49/m 14/16 +/+
66 55/f ?/9 –/? 41 41/f ?/17 +/p
1 44/m 6/9 +/+ 42 43/m 15/18 –/+
2 46/m 6/9 +/+ 43 39/m 15/19 +/u
5 66/m 7/9 –/+ 44 47/m 15/21 –/?
49 51/m 8/9 +/+ 45 45/m ?/22 +/+
21 49/m ?/10 –/? 46 39/m 20/23 –/+
4 54/f 5/10 +/+ 47 41/m 21/24 –/–
6 40/m 7/10 –/? 52 42/m 21/24 –/–
20 39/m 7/10 –/– 53 47/m 22/24 –/–
17 67/f 8/10 +/+ 48 57/f 15/25 +/–
9 72/f 9/10 +/p 54 50/f 17/25 –/?
7 52/f ?/11 +/+ 55 31/m 23/25 +/+
19 64/f ?/11 –/? 57 31/m 27/28 –/r
22 27/m ?/11 +/+ 58 32/m 29/30 –/+
23 43/m ?/11 –/r 59 55/m 27/31 –/+
3 48/f 4/11 –/+ 60 33/m 25/33 –/r
10 27/m 9/11 +/– 61 42/m 34/40 –/+
65 72/m 9/11 +/p 62 29/m 39/40 +/+
15 48/f 6/12 +/+ 63 25/m 37/42 –/+
25 46/m 10/12 –/– 64 28/m 42/46 –/?
12 38/m 11/12 –/+ 90 28/m ?/s –/+
11 27/m 12/12 –/r 82 68/f 13/s +/+
26 67/m 9/13 +/p 80 49/m 14/c +/–
13 24/f 10/13 +/+ 84 55/f ma/c –/+
24 65/f 10/13 +/p 85 40/f ma/s +/+
28 47/m 11/13 +/+ 89 50/f 34/s +/–
14 49/m 12/13 –/+ 86 28/m 37/s –/–
27 64/m 10/14 +/+ 81 29/m 38/s +/+
31 42/m 11/14 –/r 83 45/m 41/s +/+
30 52/m 12/14 +/+ 87 41/m 42/s +/–
29 45/m 13/14 +/+ 88 37/m 45/s +/+
50 72/f ?/15 +/p
—————————————————————————————————————
Sources of Information
Local medical officials told us that
hospital and public health records
of the epidemic had been con-
fiscated by the KGB. We never-
theless were able to assemble
detailed information on many
patients from a variety of sources.
(i) An administrative list giving
names, birth years, and residence
addresses of 68 people who died,
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
compiled from KGB records and
used by the Russian government to
compensate families of the
deceased (18). Comparison with
other sources of information,
including those listed below,
indicates that the administrative list
may include most or all of those
who died of anthrax. (ii)
Household interviews with
relatives and friends of 43 people
on the administrative list and with
9 survivors or their relatives (or
both). The interviews (directed by
J.G.) were designed to identify the
workplaces and other whereabouts
of patients before their illness. (iii)
Grave markers, giving names and
dates of birth and death, that we
inspected in the cemetery sector set
aside for anthrax victims. These
include 61 markers with names
that are also on the administrative
list and 5 with illegible or missing
name plates. (iv) Pathologists’
notes regarding 42 autopsies that
resulted in a diagnosis of anthrax
(14-17). All but 1 of the 42, an
unidentified man, are on the
administrative list. The notes
include name, age, and dates of
onset, admission, death and
autopsy. (v) Various hospital lists,
with names, residence addresses,
and , in some cases, workplaces or
diagnoses (or both) of
approximately 110 patients who
were apparently screened for
anthrax, 48 of whom are indicated
to have died. Of the latter, 46 are
on the administrative list. (vi) Full
clinical case histories of 5
survivors hospitalized in May
1979.
Current street and regional
maps were purchased in
Sverdlovsk, which is known again
by its prerevolutionary name of
Ekaterinburg. The city is the seat
of an administrative region, or
oblast, named Sverdlovskaya. The
city itself is divided among a
number of districts, or rayon, the
most southerly being Chkalovskiy
rayon. A satellite photograph of
the city taken 31 August 1988 was
purchased from SPOT Image
Corporation (Reston, Virginia).
Archived meteorological data
from the city’s Koltsovo airport
were obtained from the National
Center for Atmospheric Research
(Boulder, Colorado).
Fig 2. Probable locations of patients
when exposed. The part of the city
shown in the photograph is enclosed
by a rectangle in the inset. Case
numbers, in red, correspond to those
in Table 1 and indicate probable
daytime locations of patients during the
period 2 to 6 April 1979. Of the 66
patients mapped as explained in the
text, 62 mapped in the area shown.
This distribution may be somewhat
biased against residence locations,
because daytime workers not on
vacation who both resided and worked
in the high-risk zone are mapped at
their workplaces. Proceeding from
north to south, Compound 19,
Compound 32, and the ceramics
factory are outlined in yellow. The five
patients residing in Compound 32 are
mapped at their apartments. Within
the compound, the placement of an
additional, part-time resident and of the
five reservists is arbitrary, as is that of
the five residents and a nonresident
employee in Compound 19. Patients
known to have worked in the ceramics
pipe shop are mapped in the eastern
part of the factory area, where the pipe
shop is located. Calculated contours
of constant dosage are shown in black.
Approximately 7000 people lived in the
area bounded by the outermost
contour of constant dosage,
Compound 32, and the ceramics
factory. The terrain slopes gently
downward by about 40 m from
Compound 19 to the ceramics factory.
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
Case Data
Table 1 presents information on 66
patients who died and 11 who
survived. The fatalities include the
unidentified man and all people
named on the administrative list,
except for three patients for whom
recent reexamination of preserved
autopsy specimens does not
support a diagnosis of anthrax (19).
For survivors, diagnoses of anthrax
are supported by clinical case
histories or hospital lists or both
and by household interviews.
Overall, 55 of the 77 tabulated
patients are men, whose mean age
was 42. The mean age for women
was 55. No man was younger than
24, and only two women, aged 24
and 32, were under 40. Recorded
onsets span a period of nearly 6
weeks, 4 April to 15 May, with a
mean time between onset and
death of 3 days (Table 1 and Fig.
1). Approximately 60% of the 33
men for whom we have relevant
information were described as
moderate or heavy smokers and
nearly half as moderate or heavy
drinkers. None of the women was
said to have smoked or to have
consumed alcohol more than
occasionally. Few patients were
reported to have had serious
preexisting medical conditions.
Among the 35 men whose
occupation in 1979 we could
determine, the most common
occupation was welder, accounting
for 7.
In descending order of
frequency, symptoms reported in
household interviews included
fever, dyspnea, cough, headache,
vomiting, chills, weakness,
abdominal pain and chest pain.
Two of the survivors interviewed
reported having had cutaneous
anthrax, one on the back of the
neck, the other on the shoulder.
Hospitalized patients were treated
with penicillin, cephalosporin,
chloramphenicol, anti-anthrax
globulin, corticosteroids, osmo-
regulatory solutions, and artificial
respiration. The average hospital
stay was 1 to 2 days for fatal cases
and approximately 3 weeks for
survivors. To the best of our
knowledge, no human anthrax has
occurred in the Sverdlovsk region
since 1979.
Fig. 3. Villages with animal anthrax.
Six villages where livestock died of
anthrax in April 1979 are A, Rudniy; B,
Bolshoye Sedelnikovo; C, Maloye
Sedelnikovo; D, Pervomaiskiy; E,
Kashino; and F, Abramovo. Settled
areas are shown in gray, roads in
white, lakes in blue, and calculated
contours of constant dosage in black.
Public Health Responses
Public health measures were
initially directed by an emergency
commission formed in Chkalo-
vskiy rayon, where most patients
lived and worked. On or about 10
April, overall direction was
assumed by a commission that was
constituted at oblast level and
included the USSR Deputy
Minister of Health. Military
personnel participated little if at all
in the implementation of medical
and public health measures.
Before the bacteriological
confirmation of anthrax, on 11
April (14), patients were taken to
hospitals served by the ambulance
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
or polyclinic of first contact.
Starting on 12 April, most patients
presenting with high fever or other
indications of possible anthrax or
who died at home or elsewhere of
suspected anthrax were taken to
city hospital No. 40, where
separate areas were designated for
screening suspect cases and for
treating nonsystemic cutaneous
cases, for intensive care, and for
autopsy. Bodies of those who died
were placed in coffins with
chlorinated lime and buried in a
single sector of a city cemetery.
Medical and sanitation teams
recruited from local hospitals and
factories visited homes of
suspected and confirmed cases
throughout the city, where they
conducted medical interviews,
dispensed prophylactic tetracycline
to patients’ households, disinfected
kitchens and sick rooms, and took
meat and environmental samples
for bacteriological testing. Human
anthrax is not considered
contagious, nor was there any
evidence of person-to-person
transmission. In the part of
Chkalovskiy rayon where most
patients resided, building exteriors
and trees were washed by local fire
brigades, stray dogs were shot by
police, and several previously
unpaved streets were asphalted.
Newspaper articles and posters
warned of the risk of anthrax from
consumption of uninspected meat
and contact with sick animals.
Uninspected meat in vehicles
entering the city from the south
was confiscated and burned at
highway checkpoints.
Starting in mid-April, a
voluntary immunization program
using a live nonencapsulated spore
vaccine (designated STI) was
carried out for healthy persons 18
to 55 years old served by clinics in
Chkalovskiy rayon. Posters urged
citizens to obtain “prophylactic
immunization against anthrax” at
designated times and places. Of
the 59,000 people considered
eligible, about 80% were vacci-
nated at least once.
Geographical Distribution of
Human Cases
Most of the 77 tabulated patients
lived and worked in the southern
area of the city shown in Fig. 2.
Of the 66 patients for whom we
have both residence and workplace
locations, 9 lived and regularly
worked outside of this area.
Interviews with relatives and
friends revealed that five of these
nine had attended military reserve
classes during the first week of
April 1979 at Compound 32, an
army base in the affected area.
Respondents stated and, in one
case, showed diary notes
establishing that the first day of
attendance was Monday, 2 April,
that classes began at 0830, and that
participants returned home each
evening. Assuming that the
reservists were exposed while at or
near Compound 32, this must have
occurred during the daytime in the
week of 2 April.
In order to locate the high-risk
area more precisely, we prepared a
map showing probable daytime
locations of the 66 patients during
the week of 2 April. Those with
residence or work addresses in
military compounds or attending
reserve classes were placed in the
appropriate military compound;
night workers, pensioners,
unemployed people, and
vacationers were placed at their
homes; and all other workers were
placed at their workplaces. This
mapped 57 patients in a narrow
zone approximately 4 km long,
extending from the military
microbiology facility to the
southern city limit. The remaining
nine worked outside this zone, but
three of them resided within it.
Placing the latter at their
residences gives the distribution
shown in Fig. 2, with 60 of the 66
mapped cases in the high-risk
zone, 2 cases east of it, and 4 cases
north or east of the area of the
figure. Of these six patients who
both worked and lived outside the
high-risk zone, three had
occupations (truck driver, pipe
layer, and telephone worker) that
might have taken them there, one
was temporarily working in
Chkalovskiy rayon, one was on
vacation, and inadequate
information was available for
another.
At the northern end of the high-
risk zone is the military biological
facility, Compound 19, followed to
south by Compound 32. Both
compounds include numerous
buildings, with four- and five-story
apartment houses for about 5000
people at the former and 10,000 at
the latter. The administrative list
includes five people who lived in
Compound 19 and five who lived
in Compound 32. All of the latter
resided in four adjacent apartment
buildings in the eastern part of the
compound. Interviews in Com-
pound 32 indicated that all of its
residents who died of anthrax are
on the administrative list.
Interviews were not conducted in
Compound 19.
Adjacent to Compound 32 and
extending south-southeast for
about 1.5 km is a residential
neighborhood with a 1979
population density of approxim-
ately 10,000 per square kilometer,
composed of small single-story
private houses and a few apartment
houses, shops, and schools. Just
south of this is a ceramics factory
that had about 1500 daytime
employees. Of the 18 tabulated
patients who were employees
there, 10 worked in a large
unpartitioned building where
ceramic pipe was made and which
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
had a daytime work force of about
450. The attack rate at the
ceramics factory therefore appears
to be 1 to 2%. Still farther south
are several smaller factories,
apartment buildings, private
houses, schools, and shops, beyond
which begins open countryside
with patches of woodland.
Animal Anthrax
Anthrax has been enzootic in
Sverdlovskaya oblast since before
the 1917 revolution (20). Local
officials recalled an outbreak of
anthrax among sheep and cattle
south of the city in early spring
1979. A detailed report of a
commission of veterinarians and
local officials describes the
epizootic in Abramovo, a village of
approximately 100 houses 50 km
south-southeast of Compound 19.
The report, dated 25 April 1979,
records the deaths or forced
slaughter of seven sheep and a cow
with anthrax that was confirmed by
veterinary examination. The first
such losses were of two sheep on 5
April, followed by two more on
each of the next 2 days, another on
8 April, and a cow on 10 April, all
belonging to different private
owners. These losses were
substantiated by interviews we
conducted with owners of six of
the sheep that died. Respondents
said there had been no human
anthrax in the village. During a
livestock immunization program
started on 10 April, 298 sheep
were given anti-anthrax serum or
vaccine or both. The attack rate
among sheep at Abramovo
therefore appears to have been
approximately 2%.
In addition, we obtained
veterinary reports of bacterio-
logical tests positive for anthrax in
samples from three sheep from
three farms in the village of
Kashino, one sheep from
Pervomaisky, and a cow from
Rudniy, the earliest samples being
received for testing on 6 April.
Although other documents cite the
forced slaughter of a sheep in
Rudniy on 28 March and the death
of another in Abramovo on 3
April, the earliest livestock losses
for which we have documentation
of a diagnosis of anthrax are those
in Abramovo on 5 April.
Altogether, Soviet publications
(6, 7) and the documents we
obtained cite outbreaks of anthrax
among livestock in six villages:
Rudniy, Bolshoye Sedelnikovo,
Maloye Sedelnikovo, Pervoma-
iskiy, Kashino, and Abramovo.
All six villages lie along the
extended axis of the high-risk zone
of human anthrax (Fig. 3). The
centerline of human and livestock
cases has a compass bearing of
330° ± 10°.
Meteorology
Surface (10 m) observations
reported at 3-hour intervals from
Koltsovo airport, 10 km east of the
ceramics factory, were examined
in order to identify times when the
wind direction was parallel to the
Fig. 4. Wind directions and speeds
reported from Koltsovo airport for the
period 2 to 4 April 1979. Numbers at
the downwind end of each line are
local standard times. Inner and outer
concentric circles designate wind
speeds of 2.5 and 5.0 m s
–1,
respectively. Zero wind speed was
reported for 0400 on 3 April and for
0100 and 0400 on 4 April. No data
were reported for 0700.
centerline of human and animal
cases. During the time that the
reservists who contracted anthrax
were at Compound 32, but before
the first recorded human onsets,
this occurred only on Monday, 2
April, when northerly winds from
the sector 320° to 350° were
reported throughout the period
0400 to 1900 local time (Fig. 4).
During the rest of April, winds
from this sector seldom occurred,
accounting for fewer than 2% of
reports. During the period of
northerly wind on 2 April, which
followed the passage of a cold
front, the wind speed was 4 to 6 m
s–1, the temperature –10° to –3°C,
the relative humidity 50 to 66%,
the sky cloudless, and the midday
sun 39° above the horizon. These
conditions of insolation and wind
speed indicate that the atmosphere
near the surface was of neutral
stability (21). As is consistent with
this, temperature measurements at
500 to 1000 m indicated a slightly
stable atmosphere at 0400 and
1000 hours, becoming neutral by
1600.
Discussion
We have presented evidence that
(i) most people who contracted
anthrax worked, lived, or attended
daytime military reserve classes
during the first week of April 1979
in a narrow zone, with its northern
end in a military microbiology
facility in the city and its other end
near the city limit 4 km to the
south; (ii) livestock died of anthrax
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
in villages located along the
extended axis of this same zone,
out to a distance of 50 km; (iii) a
northerly wind parallel to the high-
risk zone prevailed during most of
the day on Monday, 2 April, the
first day that the military reservists
who contracted anthrax were
within the zone; and (iv) the first
cases of human and animal anthrax
appeared 2 to 3 days thereafter.
We conclude that the outbreak
resulted from the windborne spread
of an aerosol of anthrax pathogen,
that the source was at the military
microbiology facility, and that the
escape of pathogen occurred
during the day on Monday, 2 April.
The epidemic is the largest
documented outbreak of human
inhalation anthrax.
The narrowness of the zone of
human and animal anthrax and the
infrequency of northerly winds
parallel to the zone after 2 April
suggest that most or all infections
resulted from the escape of anthrax
pathogen on that day. Owing to
the inefficiency of aerosol
deposition and resuspension (22,
23), few if any inhalatory
infections are likely to have
resulted from secondary aerosols
on subsequent days. A single date
of inhalatory infection is also
consistent with the steady decline
of onsets of fatal cases in
successive weeks of the epidemic.
Accepting 2 April as the only
date of inhalatory exposure, the
longest incubation period for fatal
cases was 43 days and the modal
incubation period was 9 to 10 days.
This is longer than the incubation
period of 2 to 6 days that has been
estimated from very limited data
for humans (24). Experiments with
nonhuman primates have shown,
however, that anthrax spores can
remain viable in the lungs for
many weeks and that the average
incubation period depends
inversely on dose, with individual
incubation periods ranging
between 2 and approximately 90
days (25, 26).
The absence of inhalation
anthrax patients younger than 24
remains unexplained. Although
nothing suggests a lack of children
or young adults in Chkalovskiy
rayon in 1979, they may have been
under-represented in the aerosol
plume. Alternatively, older people
may have been more susceptible,
which may also explain the lack of
young people in epidemics of
inhalation anthrax early in this
century in Russian rural
communities (27).
It may be asked if the
geographical distribution of cases
is consistent with the distribution
expected for an aerosol of anthrax
spores released at Compound 19
under the daytime atmospheric
conditions of 2 April 1979.
Contours of constant dosage were
calculated from a Gaussian plume
model of atmospheric dispersion,
with standard deviations given by
Briggs for neutral atmospheric
stability in open country (21), a
wind speed of 5 m s
–1, a nominal
release height of 10 m, and no limit
to vertical mixing (Figs. 2 and 3).
The aerosol is assumed to consist
of particles of diameter <5 µm, as
can be produced, for example, by a
laboratory aerosol generator (28),
and to have a negligible infectivity
decay rate (<0.001 min–1) (2) and a
deposition velocity <0.5 cm s
–1
which is insufficient to cause
appreciable reduction of dosage at
downwind distances less than 50
km (29-31). Dosage contours are
not shown closer than 300 m to the
putative source, as the dosage at
shorter distances depends
sensitively on the effective release
height of the aerosol and the
configuration of nearby buildings.
People indoors will be exposed
to the same total dosage as those
outside if filtration, deposition, and
infectivity decay of the aerosol are
negligible. The negligibility of
these factors is supported by the
absence of significant dosage
reduction in field studies of
protection afforded by tightly con-
structed buildings against an
outside spore aerosol (32).
The calculated contours of
constant dosage, like the zone of
high human and animal risk, are
long and narrow. Contours are
shown at 10, 5, and 1 X 10–8 Q
spore minutes per cubic meter (Fig.
2) and at 0.5, 0.2, and 0.1 X 10–8 Q
spore minutes per cubic meter (Fig.
3), where Q is the number of
spores released as aerosol at the
source. The number of spores
inhaled is the dosage multiplied by
the breathing rate. On the
innermost contour of Fig. 2, for
example, a person breathing 0.03
m3 min–1, as for a man engaged in
light work (33), would inhale 3 X
10–9 Q spores.
The calculated dosage at
Abramova is more than an order of
magnitude lower than that at the
ceramics factory. This suggests
that sheep, reported to be more
susceptible to inhalation anthrax
than are monkeys (34), are also
more susceptible than humans.
It has been suggested that if
Compound 19 was the source,
there would have been many more
cases in its close vicinity than
farther downwind (13). This
expectation may be misleading, for
as a cloud moves downwind it also
widens. The total crosswind-
integrated dosage will therefore de-
crease more slowly with distance
than does the dosage along the
centerline. In the present case,
whereas the calculated centerline
dosage decreases by a factor of 40
between 0.3 and 3 km downwind,
the crosswind-integrated dosage
decreases by a factor of only 4.
Depending on the dose-response
relation, the crosswind-integrated
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
attack rate may decrease even
more slowly than this.
Considering, in addition, the lack
of information regarding the exact
locations of people in Compounds
19 and 32 at the time of exposure,
the distribution of cases is not
inconsistent with a source at
Compound 19.
More detailed comparison of the
geographical distribution of cases
with the calculated distribution of
dosage would require knowledge
of the precise locations of indi-
viduals in relation to the plume, the
number of spores released as
aerosol, and the relation between
dosage and response for the
particular spore preparation,
aerosol, and population at risk.
By far the largest reported study
of the dose-response relation for
inhalation anthrax in primates used
1236 cynomolgus monkeys
exposed to an aerosol of the
Vollum 1B strain of B. anthracis
(26, 35). This provided data that,
when fitted to a log-normal
distribution of susceptibility to
infection, gave a median lethal
dose (LD50) of 4100 spores and a
slope of 0.7 probits per log dose
(26, 36). This LD50 may be com-
pared with an LD50 of 2500 spores
obtained in an experiment done
under identical conditions with 200
rhesus monkeys (35) and with a
U.S. Defense Department estimate
that the LD50 for humans is
between 8000 and 10,000 spores
(8). For a log-normal distribution
with LD50 = 8000 and slope = 0.7,
the dose causing 2% fatalities, as
recorded at the ceramics pipe shop,
approximately 2.8 km downwind
of the source, is nine spores.
According to the Gaussian plume
model we have used, this dose
would be inhaled by individuals
breathing 0.03 m3 min–1 at the pipe
shop if the aerosol released at the
source contained 4 X 109 spores. In
contrast, a release 150 times larger
is estimated if the calculation is
based on an LD50 of 4.5 X 104
spores, which has been obtained
for rhesus monkeys by other
investigators (37), and if it is
assumed that spores act
independently in pathogenesis and
that all individuals are equally
susceptible (38). This estimate
would be lowered if allowance
were made for nonuniform sus-
ceptibility. If these divergent
estimates bracket the actual value,
the weight (39) of spores released
as aerosol could have been as little
as a few milligrams or as much as
nearly a gram.
In sum, the narrow zone of
human and animal anthrax cases
extending downwind from
Compound 19 shows that the out-
break resulted from an aerosol that
originated there. It remains to be
learned what activities were being
conducted at the compound and
what caused the release of the
pathogen.
M. Meselson is in the Department of
Molecular and Cellular Biology, Harvard
University, Cambridge, MA 02138, USA. J.
Guillemin is in the Department of Sociology,
Boston College, Chestnut Hill, MA 02167,
USA. M. Hugh-Jones is in the School of
Veterinary Medicine, Louisiana State
University, Baton Rouge, LA 70803, USA.
A. Langmuir is in the School of Hygiene and
Public Health, Johns Hopkins University,
Baltimore, MD 21205, USA. I. Popova is in
the Social and Political Sciences Division,
Ural State University, Ekaterinburg 620083,
Russia. A. Shelokov is in the Government
Services Division, Salk Institiute, San
Antonio, TX 78228, USA. O. Yampolskaya
is in the Botkin Hospital, Moscow 125101,
Russia.
* To whom correspondence should be
addressed.
† Deceased 22 November 1993.
REFERENCES AND NOTES
1. P. S. Brachman, in Bacterial
Infections of Humans:
Epidemiology and Control, A. S.
Evans and P. S. Brachman, Eds.
(Plenum, New York, ed. 2, 1991),
pp. 75-86.
2. Health Aspects of Chemical and
Biological Weapons (World Health
Organization, Geneva, Switzerland,
1970).
3. Posev (Frankfurt) 1, 7 (1980).
4. B. Gwertzman, New York Times, 19
March 1980, p. 1.
5. I. S. Bezdenezhnykh and V. N.
Nikiforov, Zh. Mikrobiol. Epidemiol.
Immunobiol. 1980 (no. 5), 111
(1980).
6. Veterinanya 1980 (no. 10), 3 (1980).
7. Chelovek I Zakon 117 (no. 9), 70
(1980).
8. Defense Intelligence Agency, “Soviet
biological warfare threat,” DST-
1610F-057-86 (U.S. Department of
Defense, Washington, DC, 1986);
L.H. Gelb, New York Times
Magazine, 29 November 1981, p.
31; J.P. Robinson, Arms Control 3,
41 (1982).
9. R. J. Smith and P. J. Hilts,
Washington Post, 13 April 1988, p.
1; M. Meselson, Fed. Am. Sci.
Public Interest Rep. 41, 1
(September 1988); I. S.
Bezdenezhnykh, P. N. Burgasov,
V. N. Nikiforov, unpublished
manuscript, “An Epidemiological
Analysis of an Outbreak of Anthrax
in Sverdlovsk” (U.S. Department of
State Language Service, translation
126894, September 1988).
10. N. Zhenova, Literaturnaya Gazeta
(Moscow), 22 August 1990, p. 12;
ibid., 2 October 1991, p. 6; S.
Parfenov, Rodina (Moscow), May
1990, p. 21; Pashkov, Isvestiya
(Moscow), 11 November 1991, p.
8; V. Chelikov, Komsomolskaya
Pravda (Moscow), 20 November
1991, p. 4.
11. N. Zhenova, Literaturnaya Gazeta
(Moscow), 13 November 1991, p.
2.
12. D. Muratov, Yu Sorokin, V. Fronin,
Komsomolskaya Pravda
(Moscow), 27 May 1992, p. 2.
13. L. Chernenko, Rossiyskiye Vesti
(Moscow), 22 September 1992, p.
2.
14. A.A. Abramova and L.M. Grinberg,
Arkh. Patol. 55 (no. 1), 12 (1993).
15. _______, ibid., p. 18.
16. L.M. Grinberg and A.A. Abramova,
ibid., p. 23.
17. A.A. Abramova, L.M. Grinberg,
O.V. Yampolskaya, D.H. Walker,
Proc. Natl. Acad. Sci. U.S.A. 90,
2291 (1993).
18. Russian Federation, statute 2667-1
(4 April 1992).
19. L. Grinberg, personal commun-
ication.
Reprinted with permission of: Science 1994, Volume 266, Pages 1202-1208
Official publication of the American Association for the Advancement of Science
20. V.M. Popugaylo, R.P. Sukhanova,
M.I. Kukhto, in Current Problems
of Anthrax Prophylaxis in the
USSR (Moscow, 1974), p. 50.
21. S.R. Hanna, G.A. Briggs, R.P.
Hosker, Handbook on
Atmospheric Diffusion (U.S.
Department of Energy Report No.
DOE/TIC-11223, Washington, DC,
1982).
22. D.E. Davids and A.R. Lejeune,
Secondary Aerosol Hazard in the
Field (Defence Research Estab-
lishment Suffield Report No. 321,
Ralston, Alberta, Canada, 1981).
23. A. Birenzvige, Inhalation Hazard
from Reaerosolized Biological
Agents: A Review (U.S. Army
Chemical Research, Development
of Engineering Center Report No.
TR-413, Aberdeen, MD, 1992).
24. P.S. Brachman, S.A. Plotkin, F.H.
Bumford, M.M. Atchison, Am. J.
Hyg. 72, 6 (1960).
25. D.W. Henderson, S. Peacock, F.C.
Belton, J. Hyg. 54, 28 (1956); C.A.
Gleiser, C.C. Berdjis, H.A.
Hartman, W.S. Gochenour, Br. J.
Exp. Pathol. 44, 416 (1963).
26. H.N. Glassman, Bacteriol. Rev. 30,
657 (1966).
27. A.V. El’kina, Zh. Mikrobiol.
Epidemiol. Immunobiol. 1971 (no.
9), 112 (1971).
28. J.V. Jemski and G.B. Phillips, in
Methods of Animal Experi-
mentation, W.I. Gay, Ed. (Acad-
emic Press, New York, 1965), pp.
273-341.
29. A.C. Chamberlain, Proc. R. Soc.
London A 296, 45 (1967).
30. T.W.Horst, in Atmospheric Sulfur
Deposition, D.S. Shriner, C.R.
Richmond, S.E. Lindberg, Eds.
(Ann Arbor Science, Ann Arbor,
MI, 1980), pp. 275-283.
31. A. Bovallius and P. Anas, in
Proceedings of the 1
st Inter-
national Conference on Aero-
biology, International Association
for Aerobiology, Munich, West
Germany, 13 to 15 August 1978,
A.W. Frankland, E. Stix, H.
Ziegler, Eds. (Erich Schmidt,
Berlin, 1980), pp. 227-231.
32. G.A. Cristy and C.V. Chester,
Emergency Protection Against
Aerosols (Report No. ORNL-5519,
Oak Ridge National Laboratory,
Oak Ridge, TN, July 1981).
33. D.S. Ditmer and R.M. Grebe, Eds.
Handbook of Respiration
(Saunders, Philadelphia, 1958).
34. G.A.Young, M.R. Zelle, R.E.
Lincoln, J. Infect. Dis. 79, 233
(1946).
35. J.V. Jemski, personal commun-
ication.
36. C.I. Bliss, Ann. Appl. Biol. 22, 134
(1935).
37. H.A. Druett, D.W. Henderson, L.
Packman, S.J. Peacock, J. Hyg.
51, 359 (1953).
38. H.A. Druett, Nature 170, 288
(1952).
39. R. Scherrer and V.E. Shull, Can. J.
Microbiol. 33, 304 (1987).
40. We thank A.V. Yablokov,
Counsellor to the President of
Russia for Ecology and Health, for
letters of introduction; Ural State
University and its then rector, P.E.
Suetin, for inviting us to
Ekaterinburg; S.F. Borisov, V.A.
Shchepetkin, and A.P. Tiutiunnik
for assistance and advice;
members of the Ekaterinburg
medical community and the
Sverdlovskaya Oblast Sanitary
Epidemiological Service for
discussions, notes, and
documents; interview respondents
for their cooperation; P.N.
Burgasov, USSR Deputy Minister
of Health at the time of the
outbreak, for documents regarding
livestock deaths; and People’s
Deputy L.P. Mishustina for the
administrative list of those who
died. I.V. Belaeva assisted with
interviews. We also thank B.
Ring, W.H. Bossert, P.J.M.
Cannone, M.T. Collins, S.R.
Hanna, J.V. Jemski, D. Joseph,
H.F. Judson, M.M. Kaplan, J.
Medema, C.R. Replogle, R.
Stafford, J.H. Steele, and E.D.
Sverdlov. Supported by grants to
M.M. from the John D. and
Catherine T. MacArthur Found-
ation and the Carnegie
Corporation of New York. This
article is dedicated to Alexander
Langmuir.
