CHAPTER III
Common Respiratory Diseases
Philip E. Sartwell, M. D. 1
The most important military diseases in terms of total morbidity, now that the intestinal and arthropod borne diseases have been largely controlled, is the group of acute minor respiratory tract; infections generically termed common respiratory diseases. This statement does not necessarily apply to operations in tropical areas, but it is demonstrably true of forces in the United States and other parts of the Temperate Zone under usual conditions. It can be made with even greater emphasis during periods of mobilization. While the average period of disability or hospitalization is relatively short (estimated at 6.7 days in 1945), nevertheless the loss ineffective manpower is very large. With the inclusion of influenza, these diseases were responsible for about 4 million admissions to hospital or quarters during World War It or 22percent of admissions for all causes.
In comparison with the interest taken in malaria, typhus, hepatitis, and influenza, the minor undifferentiated respiratory infections attracted little attention during the war, less, at any rate, than their prevalence would seem to justify. This is perhaps due to the ill-defined character of the group and to the probability that it includes several diseases of distinct but undetermined etiology. Furthermore, it is well known that intensive programs of field and laboratory research oil the common cold in the past have yielded comparatively little information of value; this tended to discourage further investigation when there were other fields which offered much better hope of progress. Since the war, reports of the isolation and cultivation of a common-cold virus in the embryonated hen`s egg have appeared and are certain to stimulate advances in the field.
Because of the limited amount of investigation undertaken during World War II and the present inadequacy of control methods, the major part of this chapter will be devoted to epidemiologic features of common respiratory dis ease as revealed by analysis of routinely collected statistics. A brief section will deal with the research oil the etiology, of these diseases which was done during the war. Control measures that have been studied or recommended will be summarized, and some of the unsolved problems on which further work might profitably be done will be mentioned.
1 The author gratefully acknowledges the assistance of Col. Fratis L. Duff, MC, USAF, insecuring data and reports--both statistical and narrative-for this study. In addition, Colonel Duff`s investigation of respiratory diseasemorbidityof Army divisions from the time of their activation (p. 69) has been utilized with his permission.
58
Credit for much of the effort toward the study and control of respiratory diseases during World War II should go to Brig. Gen. James Stevens Simmons, Chief, Preventive Medicine Service, Office of the Surgeon General, and his staff. It was General Simmons ‘policy to maintain a close and continuing scrutiny of morbidity rates and to initiate a special epidemiologic inquiry at posts or commands from which unusually high rates were reported. The members of the Army Epidemiological Board and its Commission on Acute Respiratory Disease sunder the direction of Dr. John H. Dingle were responsible for a large share of the research in this field.
REPORTING OF COMMON RESPIRATORY DISEASES
A definition of terms is important at the outset of this discussion. The diagnoses to be included in the expression "common respiratory diseases" are determined by Army reporting practices. Since the chief reliable source of information on incidence is the weekly statistical health report, it is necessary to accept the grouping used in that report, which is satisfactory in most respects. The health report provides for the inclusion of the acute phases of the following: Coryza, rhinitis,nasopharyngitis, tonsillitis, pharyngitis, laryngitis, and bronchitis. It is recognized that, while the etiology of none of these conditions is known, anatomical differentiations are probably of little importance in their etiology or epidemiology and. that all acute inflammatory conditions of the upper respiratory tract, exclusive of sinusitis, are best grouped together for study. Respiratory diseases with a specified etiology (as, for instance, streptococcal or diphtheritic diseases) were not included if the etiology was known and stated; there are undoubtedly many which failed to be recognized.
One exception to the statement that diseases of known etiology were separately reported is influenza. Influenza was reported separately on the statistical health report but was tabulated together with common respiratory diseases. Several points which are pertinent to this inclusion may be mentioned. First, it has become evident in recent years that in many instances it is impossible to differentiate influenza from other respiratory infections on a clinical basis. Second, laboratory tests for the recognition of influenza require much time and can be performed on so few cases that little or nothing can be done to improve the accuracy of routine morbidity reporting. Third, probably in recognition of these two facts, influenza is not reported as such from most Army installations except when a substantial epidemic is known to be in progress.
It is thus impossible to obtain any reliable estimate of how large a share influenza plays in the respiratory disease problem. Intensive studies will be necessary to determine this point. A number of recent studies have suggested that sporadic or endemic cases of virus influenza are more common in military populations than was once supposed.
In this chapter, it is pointed out that influenza is known to have been
59
prevalent among troops in the United States shortly before and during World War II at the following times: December 1940 to February 1941, December 1943 to January 1944, and in December 1945. The third of these epidemics was much less extensive than the first two. Overseas, United States troops experienced epidemics at about the same times, although the impact of these epidemics varied in different areas and, especially in tropical areas, was generally lighter than in the United States.
Evidence suggesting that the popularity of influenza as a diagnosis may have been waning during the war period is found in data secured from analysis of individual medical records. In1942, 6.8 percent of cases of common respiratory diseases (including influenza) among troops ill the United States were, diagnosed as influenza; in 1943, the percentage for this diagnosis fell to5.9; in 1944, to 3.5; and in 1945, to 1.6.
This sharp downward trend is not, consistent with what is known of the prevalence of influenza over the period, which would lead one to expect 1943 to be the high year and 1945 the next highest. It is more in keeping with a progressive decrease in willingness to make a diagnosis of influenza in the absence of specific evidence of the disease-evidence which can be obtained in only a very small proportion of cases.
Although estimates will vary widely for different times and places, it is probable that only a minority of the respiratory infections seen in military populations are caused by a known agent(if one excludes the common-cold viruses recently described, the frequency of which is as yet wholly unknown). Furthermore, an unknown share, of these illnesses are probably not infections at all but may be responses to irritant agents (as an example, hay fever) or local manifestations of systemic conditions, such as reactions to irnmunizing agents.
Certain diseases, as they have come to be better recognized, have been excluded from the common respiratory diseases group. Thus, in March 1942, a circular letter from The Surgeon General described the character of primary atypical pneumonia, which had recently been observed to be a rather important disease, in military life, and indicated that it was to be reported separately.2 The 18 October 1943 revision of the statistical health report included a space for this diagnosis. The effect of this change on reported incidence of common respiratory diseases cannot be evaluated, as such a change is usually gradual. It is likely that a considerable number of cases once listed under common respiratory diseases later came to be termed atypical pneumonia and hence disappeared from the category discussed in this chapter. Some deletions from the group ill the latter years of the war may have resulted from the increased emphasis on separate reporting of streptococcal sore throat. Provision for separate reporting of streptococcal sore throat was included in the October 1943 revision of the statistical health report. In September 1944,
2 Circular Letter No. 19, Office of the Surgeon General, U. S. Army, 2 Mar. 1942, subject: Primary Atypical Pneumonia, Etiology Unknown.
60
the statement was made that "this diagnosis includes cases of tonsillitis or pharyngitis known or suspected to be caused by the beta hemolytic streptococcus." 3 Such changes in reporting practices are in part, at least, responsible for a shift in the proportion of different diagnoses on the individual medical records.
There are two possible sources of information on the incidence of respiratory diseases during the war period. One is the statistical health report, a summary report which was forwarded weekly by each post and major unit to The Surgeon General. This was a working report, used by the Preventive Medicine Service, Office of the Surgeon General, and by other commands as the basis for epidemiologic investigations, institution of special control measures, and periodic appraisal of the Army`s health. The statistical health report included the following data which are pertinent to the subject of this chapter: (1) The average military strength of the post or command, (2) the number of admissions for common respiratory diseases, and (3) the number of admissions for influenza. These reports were consolidated in the Office of the Surgeon General to obtain service command, overseas theater, total United States and overseas, and total Army rates. They were also consolidated temporally to obtain monthly and annual rates (a month being either a 4- or 5-week period, depending on the number of Fridays which fall within the calendar month). All rates, whether for a period of 1 week, 1 month, or 1 year, were converted to an annual basis; that is, they represent the number of admissions per 1,000 troops that would have occurred lead the rate in the particular time period studied continued over a full year.
The other source of information is the individual medical record, which was completed for each soldier after discharge from hospital and forwarded to the Office of the Surgeon General. These records constitute the most reliable data on Army morbidity. However, because of the magnitude of the task of processing these records, the required detailed data from this source were not available at the time this chapter was prepared, and some of the details (for example, rates by week) are not obtainable from this source. Therefore, the statistical health report is utilized as the basis for routine analyses, employing tabulations of the individual medical record only in special studies.
An attempt was made to estimate the comparability of common respiratory disease frequencies in the statistical health report and the individual medical record. For this comparison, influenza admissions were added to the individual medical record totals. It was found that in the United States and some of the overseas areas the agreement was fairly good but that certain areas, notably the European theater and the Pacific, showed large disparities in some or all of the war years. Within the United States, rates shown in statistical health reports were uniformly, a little higher, but in none of the war years did the difference exceed 10 percent, as seen in table 15. It should be noted that the statistical health report rates consisted of incidence (total cases), whereas
3 War Department Technical Bulletin No. 92, 15Sept. 1944.
61
the individual medical record rates are admissions. It is to be expected, therefore, that the former would be higher-the amount of the statistical health report excess depending upon the extent of secondary diagnoses.
A comparison of monthly rates computed from the two sources for troops in the United States during 1942 and 1944 showed close parallelism, and the excess in statistical health report rates was chiefly evident in the first, half of both years. It is concluded that rates based on statistical health report data are a satisfactory index to the incidence of common respiratory disease, at least for troops in the United States.
Only illnesses of a certain degree of severity are reported in military practice. If the soldier who attends sick call is not considered ill enough to be taken off duty and admitted to hospital or quarters, he goes unreported. Criteria for taking men off duty varied widely, depending on the policy of the post or command surgeon and the dispensary surgeon, as well as type of duty and facilities for care. Administrative policies with respect to placing men on quarter’s status also affect the admission figures.
Certain variations in the rates may be considered as "artifacts" in the sense that then do not reflect true differences in incidence. An example of these is the usual marked drop in admissions during weekend periods, at least in situations where weekend leaves are obtainable. Most of these artifacts do not greatly affect the data used in this study, but the tendency for low rates or showing up in the normal seasonal increase at the Christmas and New Year’s Holidays needs to be kept in mind.
These remarks indicate that the disease group under discussion is a very vague group, subject to large irregularities in recognition and reporting. While this is true, it must be emphasized that few sources of information concerning mass behavior of the group approach the accuracy possessed by military records. The reasons for this are inherent in the availability of medical care, routine collection of data, and continuous enforcement of relatively simple rules for uniform reporting which the armed services provide.
62
EPIDEMIOLOGY
Epidemiologic information about common respiratory disease must be characterized as descriptive epidemiology; that is, certain facts are known about the selection of the disease as to time, place, and person without there being any real comprehension of the reasons underlying that selection. The existing information is of interest because it affords clues which, in the future, may point the way toward a more basic understanding and because it makes possible the prediction of and suggests ways of avoidance of periods of lngllmorbidity.
The time distribution of common respiratory diseases will first be considered. The period covered by this history was preceded by 2 years (1940-41) of rather high morbidity; 1942 and 1943 were years of moderately high rates, while in 1944 and 1945 the rates were quite low, as seen in table 16.
It will be noted that for troops in the United States there was a sharp downward trend in rates over this period, while for troops overseas no trend was evident, although the rates for 1945 were lower than rates for any previous year. The United States trend is at least partially explained by the recruit epidemic phenomenon which is discussed later. The proportion of recruits comprising the United States military population declined progressively during the war. As there were fewer no recruits (in the sense of men with less than 6 months` training) overseas, this trend wouldn’t be expected to be manifest in the theaters of operations.
It is not thought that influenza had a very great effect on the annual rates except in 1941 and, to a lesser extent, in 1943. Its influence on overseas morbidity was probably much less than on United States rates.
It is well known that there are large seasonal fluctuations in common respiratory disease in the Temperate Zone, but seldom have careful statistical measurements of this phenomenon been made. The "average" seasonal cycle of rates based on monthly data for the period 1924-47 has been computed. 4 Through a curve-fitting process, it was possible to construct a synthetic annual
4 Health of the Army, June and September 1947.
63
cycle which agreed closely with the average seasonal curve after smoothing and correcting forth influence of end-of-month paydays and principal holidays. This artificial cycle is the sum of three terms: The mean; a sine curve of a 1-year period with maximum in February; and a sine curve of much smaller amplitude, a one-half year period, with maxima in February and August. The highest of the weekly smoothed average rates, which occurred in mid-February, was 352;the lowest, in mid-July, was 81.5. Thus, on the average, over the United States the midwinter rate was somewhat over four times as high as the midsummer rate.
Viewing the month-by-month experience of the nine service commands, one is impressed with the essential agreement in the chronologic pattern over the entire United States. While the magnitude of rates differed between service commands and the seasonal curve for each of the four war. Years differed from all the other years, the pattern of rates ill a particular year tended to be alike all over the country, as though the "epidemiologic unit" for some of the chief components of this disease group were the Nation or perhaps even the continent, as it was for meningococcal meningitis. This may be explained in part by the influenza epidemics, which were clearly nationwide, and the proportion of new recruits in each service command rose and fell in the same fashion at the same times. There is no evidence that there were important regional influences.
Seasonal trends in overseas commands are a result of geographic location and will be discussed later (p. 65).
Chassan made a study of incidence by employing monthly respiratory disease admission rates for troops in the United States over a 40-year period from 1906 to 1946.5 He found a high correlation between rates for successive months, the coefficient of correlation averaging 0.84 and ranging from 0.71 for the December-January pair of months tip to 0.89 for the August-September pair. This means, for example, that, if in a particular year the August rate is high, the September rate is likely to be proportionately high, and vice versa. The correlation is still evident when pairs of months separated by a considerable interval of time are taken; thus, the coefficient of correlation remains positive and larger than 0.5 when the interval between the months being compared is up to 9 months. Seasonal changes do not influence the, coefficient of correlation.
As has been indicated, the most powerful factor determining common respiratory disease rates is the proportion of recruits in the command. With this in mind, geographic differences in rates may be described. Within the United States, the comparison among the nine service commands shows a general tendency toward higher rates in more northerly areas, as shown in table 17 and chart 7. The service commands, arranged in descending order of rates for the war period, are as follows: Sixth, Seventh, Fifth, Second, Third, First, Fourth, Ninth, and Eighth. This cannot be attributed to the recruit factor, since a somewhat higher proportionate recruit strength was present in the southern parts of the country owing to more favorable climatic conditions for
5 Chassan,J. B.: The Autocorrelation Approach to the Analysis of the Incidence of Communicable Diseases. Human Biol.20: 90-1C8, May 1948.
64
65
training. Examination of the seasonal curves of incidence indicates that the magnitude of the seasonal rise and fall was greater in those service commands which had the highest rates.
The over-all experience of the overseas areas will next be considered. The comparative rates should not be given great weight, since troop strengths varied widely in different theaters. The distribution of personnel within the areas was not constant. For example, no United States troops were in the Mediterranean theater until late in 1942. During most of1943, all United States forces in this theater were in North Africa, while in the succeeding years they moved into Italy where environmental conditions were quite different. Thus, it is apparent that the theater designation is only an approximate and inconstant description of the geographic area of service. Certain generalizations can, however, beamed. The average rate of respiratory diseases (203) was higher in the Alaskan Department than in any other major region. The North American area(with an average rate of 193) was next, succeeded by the Middle East(184), China-Burma-India(151), Mediterranean (including North Africa) (145), and European (137)theaters; and Southwest Pacific (111), Latin American (including Antilles Department)(103), and the Pacific Ocean (81) areas in that Order. There are very marked differences in the extent of seasonal variations in these areas which do not entirely parallel the rates(chart 8).Seasonal variation was most extreme in the European theater, was moderate in the Alaskan, North American, Mediterranean, and Middle East regions; and was very small or negligible in the China-Burma-India theater, Latin America, Southwest Pacific, and Pacific Ocean areas. One may say, generally, that the more northerly regions lead both the higher rates of admission and the more marked seasonal rises and falls. Rates were low and constant ill parts of the world with tropical and equable climates. For example, in
66
67
the Latin American area, the average rate was 103 per 1,000 per year over the 4-year period. Expressed as a proportion of total disease admissions, common respiratory diseases and influenza ranged downward from 37 percent in Alaska and the North American area to 13percent in the Southwest Pacific Area.
It is commonplace to hear that the climate in a particular locality is conducive to minor respiratory diseases. A systematic analysis of rates at individual posts would have to take into consideration all the following factors: The length of service of the troops; rate of turnover; type of duty; age composition; and the prevalence of recognized epidemics of influenza, streptococcal disease, or other specific agents. Much of this information is not available, and such studies, therefore, will not be attempted here. It may be stated, however, that stations within the Sixth and Seventh Service Commands, particularly in the States of Colorado and Wyoming, consistently experienced comparatively high rates of admission for common respiratory diseases. There are several reasons which suggest that these high rates reflect in part an excessive prevalence of streptococcal disease at certain stations.
The influence of race,sex, and age cannot, unfortunately, be studied in detail owing to thedearthof information available at the present time. There have been three studies which have suggested that Negro troops experience a lower rate of admission than do white troops. In the first of these, Gordon found that rates in the European theater during World War II were consistently lower for Negro than for white troops.6 The Commission oil Acute Respiratory Diseases found a similar difference in its intensive studies at Fort Bragg, N. C., but could not exclude the possibility that it was due to differences in policy with respect to placing men under treatment. The author encountered substantial differences at all seasons between the admission rates of white and Negro recruits in his investigation at Fort Dix, N. J., in 1947-50. 7Both groups showed large epidemics in units recruited and assembled during the winter months, but the magnitude of these rates was usually more than twice as high during the first few weeks of service for white troops as for Negro troops. No reasons based on administrative policy or factors of morale could be discovered to explain the difference.
With respect to sex, a limited number of observations on Women`s Army Corps groups have indicated that admissions for common respiratory diseases, as for numerous other causes, are more frequent than among male troops. Thus, during the period from June 1944 to December 1945, inclusive, the admission rate for common respiratory diseases and influenza among Women`s Army Corps personnel in the, United States was some 60 percent higher than that for all Army personnel, though the pneumonia admission rate was slightly lower than that for all Army personnel.8 This is consistent with experience reported by a number of industries and the findings of a majority of com-
6 Gordon, J. E.: A History of Preventive Medicine in the European Theater of Operations, U. S. Army, 1941-45 [Official record.]
7Sartwell, P. E.: Common Respiratory Disease in Recruits. Am. J. Hyg. 53: 224-235, March 1951.
8 Enlistment, Health, and Discharge of the WAC. Bull. U. S. Army M. Dept. 6: 276-287, September 1946.
68
munity surveys of illness morbidity. It maybe doubted whether the incidence of respiratory infections is higher among females under military circumstances than among males. The difference may represent a greater tendency for Women`s Army Corps personnel to report for medical treatment or a greater likelihood of their being admitted to hospital or quarters by the medical officer when they seek such care.
It has not been possible to secure any data with respect to admissions for common respiratory diseases by age. However, some studies have indicated that, over the span of ages represented by military organizations, the incidence of such illnesses does not vary much with age. As these illnesses do not appear to produce any prolonged immunity, this is the expected pattern.
INCIDENCE AMONG RECRUITS
The phenomenon of high incidence of respiratory infections among recruits has been universally noted and commented upon by students of military medicine. The most detailed observations made during World War II were those of the Commission on Acute Respiratory Diseases of the Army Epidemiological Board. 9 A somewhat similar study was conducted at Fort Dix by the author over the period 1947-50.10 In both studies, morbidity rates specific for week of service were obtained by following units of company or battalion size from the time of their organization. The studies cited have, in general, shown that recruits arriving on the post in winter and thrown together with other recruits in training companies usually experience, almost at once, sharp epidemics of upper respiratory diseases. In the, author’s studies, these epidemics most commonly reached their peak: in the third or fourth week of training, and incidence was down to a low level by the sixth to eighth week. Sometimes the rates go as high as 3,000 admissions per1,000 troops per year for this brief period, returning to a baseline of less than 300 per 1,000 per year at the end of the epidemic. This process went on in the absence of any epidemic conditions among seasoned personnel on the post. Among the units observed, there were few which failed to undergo an epidemic, but the observations are too limited to say how general such experiences are in different years or in other parts of the country.
Troops inducted and beginning their training at other seasons of the year were also observed. Those inducted during the summer months generally lead no epidemics, while those inducted in the spring and fall underwent a variety of experiences, seldom, however, having the explosive, outbreaks that were characteristic of units brought in the service during the winter. The Fort Dix study showed that, of troops arriving in January or February of the winters of 1947-48 and 1948-49, 12 and 15 percent, respectively, were admitted
9 (1) Commission on Acute Respiratory Diseases: Acute Respiratory Disease Among New Recruits. Am. J. Pub. Health 36: 4:39-450, May 1946. (2)Commission on Acute Respiratory Diseases: Epidemiology of Atypical Pneumonia and Acute Respiratory Disease at Fort Bragg, North Carolina. Am. J. Pub. Health 34: 335-346, April 1944.
10 See footnote 7,P. 67.
69
to hospital during their first 8 weeks of military service and that during the summer of 1949 only about 1 percent were admitted.
The following discussion is based on a study conducted under the author`s supervision by Col. Fratis L. Duff, MC. This study has provided new and confirmatory evidence regarding the effect of seasoning of troops as well as the climatic and geographic factors in common respiratory diseases.
During the period 1940-41, nearly all the Army divisions ordered to active duty were National Guard divisions. Officers and men of existing guard organizations were inducted as units, and the divisions were brought to strength by assignment of newly inducted men. Troops in these divisions consisted mainly of men who had lived in the same general area in which they received their training. In 1942, new Army of the United States and Reserve divisions were activated by assigning sufficient officers and noncommissioned officers to serve as a training cadre and their by rapidly filling the ranks with newly inducted troops. Troops in these new divisions came from all parts of the country. The period of training in the United States for both types of divisions was never less than 1 year. During the course of training, all divisions received some additional troops as replacements for mere discharged from service or reassigned to other units. However, the number of changes in composition was usually quite small in proportion to the total divisional strength. Overseas, the personnel turnover rate was considerably larger owing to combat losses, reassignments of subordinate units, and later, the return of personnel to the Zone of Interior. After 1943, all recruits spent a period of at least 2 months in training at a basic training center before receiving their first operating assignment.
Quotas for most of the divisions formed in the early part of the war were filled within a period of a few weeks, so that, early in their training the divisions consisted of a large majority of troops having approximately the same length of service, together with some seasoned men having diverse military backgrounds. This, as Colonel Duff recognized, provided opportunity for a study of common respiratory diseases in divisions, with the seasoning factor at work. Furthermore, since divisions began training at different calendar periods and in different parts of the country, there was hope of separating for study the epidemiologic factors of time, place, and military age. Colonel Duff was able to obtain the statistical health reports of 19 division’s acid. 1 artillery battalion covering periods of at least 1 full year from the time of organization. Some of these divisions could be followed for periods of 3 or 4 years, including long periods of overseas service and combat operations.
The information abstracted from each statistical health report included the weekly average strength and the number of admissions for common respiratory diseases, influenza, and outer acute communicable diseases; however, only the common respirator diseases and weekly average-strength figures were systematically utilized. From other sources, information was obtained as to periods of service of the division at each post or maneuver area in the United States, time of embarkation for foreign duty, and combat
70
periods while overseas. An arbitrary rule was made that the period of service of a division would be dated from the month in which the increase in strength from cadre to full or nearly full operational strength took place. In other words, for the purposes of this epidemiologic study, the time of mobilization of the division was taken as that month in which its recruit population, rather than its cadre, was assembled.
The experience of each unit during its first year of training was intensively studied. Admission rates for common respiratory disease were computed for the first 4-, 8-, and 12-week periods, for the first and second 6 months, and for the entire year (table 18). Monthly rates for the year were also computed; rates for organizations selected as typical are reproduced graphically in charts 9 and 10.
TABLE 18.-Admission rates for common respiratory diseases in selected units
Divisions mobilized in the northern half of the United States had higher rates over the first year, as a rule, than those. In the South. Of the 7 divisions trained in the States in the North, 4 had average rates of 600 or higher over the first year, while of the 12divisions in the South, there were none with a rate above 600.
The incidence curves of these 19 divisions are quite varied, but there is an underlying common pattern which may be described in general terms. In
71
the divisions for which there were several years` records available, the rate for the first winter after formation of the division was always highest, the second and third winters having successively diminishing rates. Rates were regularly higher for divisions training in the North than for those in the South, as already stated. Divisions formed in the period from May to October experienced a sharp peak in either the succeeding December or January, but there was usually only 1 epidemic month. One exception to this statement was the89th Infantry Division, which trained at Camp Carson, Colo., and experienced a peak in December1942 but continued to have rather high rates until the following June, possibly representing the incidence of streptococcal disease. Units having the highest rates were those assembled during the winter months in the Northern States. There was no tendency for high
72
rates to be associated either with periods of training in maneuver areas or with combat operations overseas.
Examples of the most extreme effect of the recruit epidemics on divisional rates are the 26th and 33d Infantry Divisions, both National Guard units, which were largely composed of men residing in the same section of the country where the divisions were assembled. The 26th was ordered into Federal service on 16 January 1941 and moved to Camp Edwards, Mass., in February 1941 (the month in which the strength rose, above 8,000). In that month, the rate was 3,415, dropping precipitously in succeeding months to 120 in July. The next winter, while again stationed in Massachusetts, the division experienced its highest rate of 372 in February. The 33d Infantry Division was ordered into Federal service on 5 March 1941 at Chicago, Ill., and shortly afterward moved to Camp Forrest, Tenn. The strength first rose to 10,000 at the end of March 1941, and in the following month the admission rate was 3,697, thereafter dropping dramatically to 324 in August. Maximum rates of about 800 were experienced in the following two winters.
The effect of this high recruit susceptibility is, of course, high respiratory disease rates in the mobilization period of wars and in troops in the United States as contrasted with those in overseas commands. It has already been pointed out that the trend of rates for troops in the United States was downward from 342 in 1940 to 116 in 1945. No direct estimate of the proportion
73
of unseasoned troops in the United States at various times has been found, but figures for the number of inductions by months have been obtained from The Adjutant General, and so it is possible to compute for any period during the war the number of men inducted during the previous 3 months. Some of these men were discharged before completing3 months` service, and a very few of them, perhaps, were assigned to overseas duties. The large majority, however, were still undergoing training in the United States 3 months after induction. Hence, this number (inductions during the past 3 months) was taken in relation to the total troop strength in the United States each month, as a rough estimate of the percentage of troops in this country with less than 3 months` service. By means of a series of scatter diagrams, the correlation between this figure and the corresponding United States common respiratory disease rate for each calendar month over a 5-year period (1941-45) was studied. In nearly all months, some correlation was evident; it was somewhat greater in the colder months. For the months of December 1943 and January 1944, incidence was higher than expected from the recruit percentage (influenza A being epidemic during this period). The correlation for the months of January and February is indicated in chart 11. This association does not of itself, of course, imply direct relationship between respiratory disease rates and mobilization but is quite consistent with the abundant evidence from other sources.
RESPIRATORYINFECTIONS ABOARD TRANSPORTS
Considerable trouble was experienced in the European theater during 1942 with the high incidence of upper respiratory infections acquired by troops on the ocean voyage from the United States. Of three units surveyed, 51 to 75 percent lead colds during a period of approximately 5 to 7 weeks including the pre embarkation period, the voyage, and the post embarkation period despite the fact that the voyage was in late summer. The frequency of colds was low prior to embarkation, increased more than tenfold while in transit, and continued high, though subsiding, after arrival. Overcrowding on shipboard, which was unavoidable at the time, was considered responsible. On the other hand, a convoy of 24transports left New York and Boston in December 1943, at the height of the influenza epidemic of that year and at a time when respiratory diseases were prevalent in the New York: staging area. Although serious difficulty was feared, this convoy which carried 63,750 troops encountered no more respiratory diseases during the voyage than would be expected at the season, in the opinion of ships ‘surgeons and investigators in the theater.
ASSOCIATION WITH OTHER DISEASES
The seasonal pattern of most of the diseases transmitted by respiratory secretions is much the same. Measles, meningococcal meningitis, mumps, and streptococcal infections are at their height during the late winter or spring; diphtheria (in the United States), in late autumn; pneumonia, usually in mid-
74
winter or late winter. Furthermore, most of these diseases show a relationship to length of service in general like that of common respiratory disease. These two factors are sufficient to account for the apparent association of common respiratory disease with other respiratory infections. It is possible that coughing, sneezing, and excessive respiratory secretions characteristic of colds facilitate the transmission of other diseases. In Army camps it was common, though by no means invariable, when common respiratory disease was epidemic, for primary atypical pneumonia and often measles, mumps, or streptococcal infection also to be prevalent. An association with meningococcal meningitis was observed in the winter of 1942-43in the United States.11 On the other hand, the sharp outbreaks of influenza in 1943 and 1945occurred at a time when the common respiratory disease rate was rather low.
11 Sartwell, P. E., and Smith, W. M.: Epidemiological Notes on Meningococcal Meningitis in the Army. Am. J. Pub. Health 34: 40-49, January 1944.
75
ETIOLOGICSTUDIES
Most of the investigations of possible etiologic agents of so-called undifferentiated respiratory diseases made during the war under Army auspices were carried out by the Commission on Acute Respiratory Diseases of the Army Epidemiological Board. This Commission, headed by Dr. Dingle, maintained its laboratories and made its field observations at Fort Bragg. The Commission’s findings have been published in a comprehensive series of papers and no attempt will be made here to do more than restate their salient features. Studies were made of the bacterial flora of the upper respiratory tract which led to the conclusion that the flora of patients hospitalized for undifferentiated acute respiratory diseases did not differ from that of healthy recruits.12 In another study, only 6 percent of the cases of mild respiratory infection was shown to be streptococcal infection by bacteriologic and immunologic studies.13There are indications that streptococcal disease was quite uncommon during the war in this region. In a group of patients with exudative pharyngitis and tonsillitis, a condition which was found to constitute some 10 percent of respiratory disease admissions, the Commission reported that 25 percent had beta hemolytic streptococci in their throats and exhibited a rise in titer of streptococcal antibodies during convalescence, another 25 percent had streptococci but did not develop antibodies, and the remaining one-half lead no streptococci. 14 The conclusion was drawn that in approximately one-fourth of this series of cases the disease was of streptococcal etiology. No means of accurate diagnosis other than cultural and serologic tests were found.
The Commission undertook a series of studies employing human volunteers in attempts to transmit, experimentally, the agents of minor respiratory illness in which no bacterial etiology could be found.15 The investigators were successful in producing illnesses of two clinical types by the use of bacteria-free filtrates of pooled nasal and pharyngeal washings. Subsequently, homologous and heterologous immunity was tested on the same volunteers. Immunity to inoculation was found in individuals receiving a filtrate which induced minor illness with all incubation period of about 5 to 6 days. Such immunity was not demonstrated in persons given the filtrate which induced a coryzalike illness with incubation period of 1 to 2 days. No cross-immunity was demonstrated.
12 Commission on Acute Respiratory Diseases: Bacteriological Findings in Undifferentiated and Other Acute Respiratory Diseases. Medicine 26: 465-484, December 1947.
13 Commission on Acute Respiratory Diseases: The Role of Lancefield Groups of Beta-Hemolytic Streptococci in Respiratory Infections. New England J. Med. 236: 157-166, January 1947.
14 Commission on Acute Respiratory Diseases: Endemic Exudative Pharyngitis and Tonsillitis. Etiology and Clinical Characteristics. J. A. M. A. 125: 1163-1169, 26 Aug. 1944.
15 Commission on Acute Respiratory Diseases: Experimental Transmission of Minor Respiratory Illness to Human Volunteers by Filter-Passing Agents. I. Demonstration of Two Types of Illness Characterized by Long and Short Incubation Periods and Different Clinical Features. J. Clin.Investigation 26: 957-973, September 1947. II.Immunity on Reinoculation With Agents From the Two Types of Minor Respiratory Illness and From Primary Atypical Pneumonia. J. Clin. Investigation 26: 974-982, September 1947.
76
The question of etiology is obviously still far from solution. As stated earlier, the possibilities must include not only infectious agents but irritant or allergic incidents, though there is little doubt that the large proportion of the cases are infections. An interesting subject of speculation is the relationship between minor infections of the upper respiratory tract and primary atypical pneumonia, itself a vaguely defined disease group. Some parallelism between the two conditions has at times been noted, though this alone should not be taken as indicating any etiologic relationship.
The diagnosis of common respiratory disease remains a matter of exclusion of recognized disease entities, such as streptococcal tonsillitis and pharyngitis, influenza, pneumonia, and the like. Much information with a bearing on common respiratory disease will be found in publications relating to these other diseases, notably those of the commissions of the Army Epidemiological Board on influenza, hemolytic streptococcal infections, and pneumonia.
PREVENTIONAND CONTROL
During the war, promising measures for the control of certain specific respiratory infections were introduced. Influenza vaccine received its first large-scale trials; chemoprophylaxis proved successful under certain conditions in the control of streptococcal and meningococcal infections; methods for air sanitation (glycol vapors, ultraviolet irradiation, and dust-suppressive measures) were tried. Both chemoprophylaxis and air sanitation were based on the demonstrated effectiveness of these agents as weapons against streptococci. The reason for mentioning them here is the possibility that they may also be effective against the agent or agents of common respiratory disease. Sulfadiazine chemoprophylaxis has not, so far as known, been subjected to any rigid test of this possibility. From the lack of evidence that it has clinical value, however, it may be presumed that (1) its prophylactic effectiveness is due to a reduction in streptococcal or other bacterial infections and (2) if such infections are not prevalent, it will not reduce respiratory morbidity. Air sanitation measures have been demonstrated to reduce the bacterial count of the air, and whatever efficacy they may possess is presumably a result of this action. The Commission on Acute Respiratory Diseases was unable to show that oiling of floors and blankets, a dust-suppressive measure, reduced the incidence of respiratory diseases at Fort Bragg during a period when bacterial infections were infrequent.16In theory, at any rate, one of the factors determining the effectiveness of air sanitation in reduction of morbidity is the extent to which the disease in question is airborne. No precise knowledge is available on the relative importance of direct or indirect contact, inhalation at close range of droplets of respiratory secretions, and true airborne infection (inhalation of minute droplet nuclei or dust particles containing an infectious
16 (1) Commission on Acute Respiratory Diseases and Commission on Airborne Infections: Study of the Effect of Oiled Floors and Bedding on the Incidence of Respiratory Disease in New Recruits. Am. J. Hyg. 43:120-144, March 1946. (2) Whayne, Tom F.: Housing. In Medical Department, United States Army, Preventive Medicine in World WarII.Volume II.EnvironmentalHygiene.Washington: U. S. Government Printing Office, 1955,pp. 27-74.
77
agent and suspended in the atmosphere of enclosed spaces). It may be that the relative importance of these various modes of transmission differs for each of the respiratory infections (fig.1).
Related to this problem is the question, to what extent is common respiratory disease transmitted in barracks, where men spend their sleeping hours, and to what extent is the disease transmitted in classrooms, mess halls, theaters, post exchanges, or other places where men congregate during the day. For the sake of completeness, still another possible source of infection needs to be mentioned; namely, foods and eating utensils contaminated by food handlers. There is little reason to regard the latter as important in this class of diseases, though it is true that occasional outbreaks of streptococcal pharyngitis have been traced to this source.
Observations of the Commission on Airborne Infections of the Army Epidemiological Board with respect to contamination of the environment with hemolytic streptococci are important to the understanding of atmospheric transfer of pathogens. Findings of this Commission demonstrated that in
78
wards and barracks the number of streptococci in bedding and floor dust is often large and that such activities as bed making and dry sweeping result in there suspension in the air of these organisms. It was also observed that oiling of floors and blankets and introduction of triethylene glycol vapors into the air had a pronounced effect in reducing the content of viable streptococci in the atmosphere of ward or barracks. Unfortunately, no conclusive evidence that these measures will reduce the prevalence of common respiratory diseases has been produced.
It has long been assumed, largely on theoretical grounds, that reduction or avoidance of overcrowding is the most important measure within reach of the military surgeon for control of these diseases. Emphasis has been placed on the allotment of barracks space to provide every man with a certain minimum floor area or cubic footage of air volume and to maintain a certain minimum space between beds. Standards for minimum crowding have been established in War Department directives. The allowable minimum floor space per man, for example, was 60 square feet before World War 11,17 but this was temporarily reduced to 40 square feet during the war except for barracks housing recruits where space allotment was 50 square feet. 18 The change in standard is significant; it is probable that during any rapid mobilization a shortage of housing facilities for troops would develop, since it is possible to expand an Army faster than new housing can be built. This means that crowding is likely to be more severe at the same time, that respiratory infections are at their peak owing to the large numbers of unseasoned recruits. To what extent crowding contributes to the difficulties is impossible to say. While the Medical Department has expressed its awareness of and opposition to the evils of crowding, it cannot be claimed that its views have been very effectively championed or upheld by command under the stress of military necessity.
Special epidemiologic investigations were made at various times during World War II when respiratory disease rates in particular posts rose to unusual levels. These investigations were conducted by post or organizational preventive medicine officers, by service command or theater headquarters, and in some instances by representatives of the Office of the Surgeon General. One investigation which will be described was carried out by representatives of the Army Epidemiological Board. Space allowance for troop housing was reduced in October 1942 from60 to 40 square feet per man. 19 The Surgeon General subsequently requested that the Board survey, conditions in Army camps resulting from the application of this order with particular reference to the current and expected incidence of acute respiratory diseases and meningococcal meningitis. Ten members of the Board and its commissions inspected 19large posts in all parts of the country during December 1942.
17 Army Regulations No. 40-205, 15 Dec. 1924.
18 (1) Army Regulations No. 40-205, 31 Dec. 1942.(2) Army Regulations No. 40-205,changes No. 2, 26 Mar. 1943.
19 (1) Letter, The Adjutant General, to Commanding Generals, Commanding Officers, et cetera, 21Oct. 1942,subject: Reduced Space Allowance at Posts, Camps, and/or Air Force Stations. (2) See footnote 18 (1).
79
Not only crowding but also heating, ventilation, messing facilities, and conditions in lavatories, post exchanges, and elsewhere were considered by this group. Crowding was found at some posts but not at others, and there. Was no particularly marked correlation between the extent of crowding and the seriousness of the respiratory disease problem. Most of the investigators felt that factors other than crowding in barracks were important, and several stated that they did not believe that crowding could be blamed for the respiratory disease rates. All, however, agreed as to the desirability of efforts to reduce crowding in barracks. At special meeting of the Board held on 29 January 1943 and attended by representatives of the Office of the Surgeon General, Army Ground Forces, Army Air Forces, Services of Supply, and the Chief of Engineers, the findings of the investigators were summarized and discussed. It was emphasized that the order reducing space allowances to 40 square feet was a command decision made on the basis of military necessity. It was suggested that a return to 60 square feet per mall in recruit reception and training centers would provide a partial solution and would automatically slow down recruitment. The Board finally adopted a resolution emphasizing the influence of crowding in barracks, mess halls, and recreation halls on the spread of meningitis, acute respiratory, and other epidemic diseases and endorsed the action of The Surgeon General in advocating a minimum of60 square feet per man. It was pointed out that this would relieve crowding not only in barracks but in mess halls, washrooms, latrines, post exchanges, and elsewhere. Following this action, the regulation on space allowance was revised on 26 March 1943 and provided minimum of 50square feet per man in reception centers, except in emergencies, and 40square feet at other stations.20
Other preventive measures which have long been taught, in military medicine and are frequently employed are the provision of adequate ventilation arrangement of bunks to permit head-to-foot sleeping, and erection of shelter halves as screens between beds. Like the spacing out of beds, these methods have received no rigid test. All are based on the assumption, which may be doubted, that transfer of respiratory infections occurs chiefly in barracks. It is most difficult to enforce corresponding measures in places of daytime congregation, particularly as the troops themselves are gregarious and tend to gather in social groups during off-duty hours. It is probable that many infections are spread on troop trains and in buses, since the frequency of travel is generally increased in military life.
One controlled study of the effect, of barracks living arrangements was made by the Commission on Acute Respiratory Diseases.21 This was a comparison of incidence rates for common respiratory diseases among recruits in barracks where double bunking was employed with those in barracks having standard single beds. Double bunking in this study meant the use of double-
20 See footnote 18 (2), p. 78.
21 Commission on Acute Respiratory Diseases: The Effect of Double-Bunking in Barracks on the Incidence of Respiratory Disease. Am. J. Hyg. 43: 65-81, January1946.
80
decked beds so spaced that the total cubic footage of airspace per man was not reduced from that prevailing in single-bunked barracks. Under these conditions, no great difference in respiratory disease incidence was seen, but the occupants of double-bunked barracks fared somewhat better, in that during an epidemic of common respiratory disease their admission rates to hospital were lower.
Numerous other factors have been incriminated as predisposing to respiratory infections. Weather has, of course, long been suspected of influencing morbidity, but such careful studies as have been made have not supported this view. Avoidance of exposure to inclement conditions has, at any rate, been a common precept during epidemic conditions. In this connection, it is of interest that both in the author`s studies and in the investigations of morbidity of divisions by Colonel Duff, the periods when troops were on maneuvers-sleeping in the open, often under severe weather conditions-were almost never marked by an increase in reported morbidity. The same, is true of periods of combat during the, winter season. It may, of course, be argued that opportunities for hospitalization of men with minor illnesses were interfered with, at such times, although this would not be, true of the maneuver periods at Fort Dix. In summary, while temporary weather disturbances and exposure of inadequately clothed troops to vicissitudes of the weather are often blamed for respiratory epidemics, there seems to be little factual foundation for this beyond the well-established facts that morbidity is greater in colder climates and there is a well-marked seasonal cycle of morbidity which is roughly inversely proportional to temperature.
While nothing can be done about the weather beyond providing proper clothing and housing and, so far as possible, locating basic training establishments in the South, there are other respiratory disease control measures which may be advocated, again on largely theoretical grounds. By analogy with results of experimental epidemiology, it would appear that communicable diseases have a better chance of spreading among troops housed in large dormitories than when the unit of accommodation is small, even though space allowances per man are equally large. This is factor which should be susceptible to study. The construction of barracks, particularly theater of operations type, was often such as to allow drafts. The temperature, control provided by the heating systems in these barracks left much to be desired, and, although measures were taken to increase the humidity, it was a common experience in winter for the atmosphere of the barracks to be very dry. All of these environmental conditions are, of course, subject to control provided the expansion of the services is not too rapid for new construction to keep pace with it. There have been arguments as to the relative healthfulness of barracks and tents in prevention of respiratory diseases without, so far as is known, any soon evidence having been produced on either side.
Another preventive measure which might theoretically be expected to reduce morbidity is the prompt recognition, diagnosis, and admission to
81
hospital of infected individuals, thus removing sources of infection from the population. During epidemic periods, however, the hospitalization of mild cases would almost certainly overcrowd the hospitals. It may be doubted whether this measure would be helpful, since it is most unlikely that all infected individuals would be removed from duty even under favorable conditions. Moreover, if the behavior of this infection resembles measles or similar diseases, many would have had an opportunity to spread the infection before hospitalization. It must be acknowledged, in any case, that the hospitalization of all cases ill enough to go off duty is preferable to treating them in quarters, both because of the reduced risk of spreading infection and the relatively better care provided in hospital.
Finally, the discussion of control measures must include health education. In technical manuals, lectures on personal hygiene, posters, training films, and other training aids, emphasis was given to the mode of transmission of respiratory infections, the principles of avoidance of overcrowding, good ventilation, covering the mouth and nose when coughing or sneezing, proper clothing, and other steps which the soldier might himself take to modify his environment. It is impossible to form any estimate of the effectiveness of this teaching in influencing the behavior of soldiers. In general, however, such instruction is to be commended.
Another factor which has been blamed for the high recruit susceptibility is the fact that all recruits receive a course of immunizations, including vaccination against smallpox, three inoculations of typhoid-paratyphoid vaccine, and three injections of tetanus toxoid. The usual policy is to administer these as soon after entry into the service as possible. There are various reasons to discount this course of immunization as an element in recruit susceptibility. One is that immunizations are given at all seasons, while recruit epidemics are infrequent in summer and epidemics in units mobilized in the autumn and presumably immunized at that time have, on a number of occasions, occurred somewhat later in the winter. Furthermore, units alerted for overseas duty have commonly received intensive mass reimmunization and immunization with additional agents such as cholera, typhus, and yellow fever vaccines without experiencing such epidemics. It is true that some reactions to immunizing agents such as typhoid vaccine are probably misdiagnosed as acute febrile respiratory infections, but this would not be sufficiently common to affect the rates materially. Still another suggested explanation for the excess incidence among recruits is the physical depletion resulting from rigorous training program. Most of the arguments advanced against immunizations and excessive exposure to the elements as factors are equally appropriate here.
Since none of these explanations of the recruit epidemic phenomenon seems adequate, a better explanation must be sought. Such an explanation can, it is believed, be found in the aggregation of large numbers of individuals coming from different environments. It seems reasonable that under ordinary civilian
82
circumstances there is relatively limited interchange of infectious agents between individuals who are in contact with one another. Living habits are such that individuals are in intimate association with a more or less constant group, although a less close association does occur with individuals encountered by chance in public vehicles, theaters, and the like. When a military organization is formed from men drawn from civilian life, it is almost certain that some of the men will be exposed for the first time in months or even years to pathogenic agents to which their immunity has waned and that they will, as a result, develop a clinical infection. Upon the termination of these clinical illnesses, it may be presumed that at least a transient immunity will remain, since some immunity mechanism is necessary for recovery from an infection. Thereafter, repeated exposures during the period when immunity is waning might suffice to reinforce that immunity without the individual suffering a second clinical attack. The fact that recruit epidemics occur only in the colder months is difficult to explain on this or any other hypothesis. It must be supposed that ecologic influences at other seasons simply do not support the development of an epidemic.
A phenomenon somewhat resembling the recruit epidemics on a smaller scale probably takes place at the beginning of a winter term in boarding schools and occasionally in other schools. Records adequate to study this are not available except in a few institutions, but those available tend to support this concept.
There is a real opportunity for research in military preventive medicine through further investigation of the various environmental factors discussed above in relation to acute respiratory diseases. The inability to attach a specific etiologic diagnosis to this group of diseases has probably been one of the chief deterrents to such studies, thought need not be if the recognized pathogens are reasonably well excluded. There are many features of Army life which would facilitate studies of this character. One possible avenue of approach to the control of respiratory diseases which has never been deliberately tried is the administrative rearrangement of programs for the training of newly inducted soldiers so as to avoid those conditions which have been shown to favor recruit epidemics, that is, the rapid aggregation of large numbers of recruits at a single post during winter months with a constant influx of new men. Admittedly this method of conducting basic training is the most efficient and economical. However, if some other method were to prevent the hospitalization of large numbers of troops, it might well prove less costly and more effective in the long run.
The argument is sometimes advanced that the seasoning process involving an attack of a respiratory disease is a necessary element of adaptation to military life. It is claimed, therefore, that attempts at prevention of recruit epidemics are unwise, since it is better for the recruit to have his seasoning behind him early in military training. As pointed out in the Fort Dix study,
83
this viewpoint is entirely defensible if one is considering diseases like measles and mumps in which the severity of attack is not thought to be influenced by dosage of the infectious agent or other factors and in which immunity is permanent. With respect to common undifferentiated respiratory diseases, there is little evidence that troops who have been through a recruit epidemic are any more adequately seasoned than those who have not. It may well be that consideration of the epidemiologic evidence now available in the planning of troop-training programs would have a definite effect in the lowering of over-all morbidity.
