CHAPTER VII
Other Statistical Studies in World War I
After the completion of the compilation and analysis of the statistics of the medical records of World War I and of the reports of the physical examination of the Selective Service registrants, there still remained several important statistical studies to be undertaken by the Surgeon General`s Office. These were:
1. A study of the relation of war casualties, including sick, nonbattle injuries, and battle injuries, to the requirements for an efficient medical and hospital service and to the replacements` requirement for the combat and service troops.
2. A long-range survey of the service efficiency and longevity of Army officers in relation to their health and physical measurements, both at the time of their admission to the military service and during their service.
3. The preparation of standard tables of height, weight, and chest measurements, based upon the military experience both during periods of mobilization and peacetime.
Gen. Merritte W. Ireland, then Surgeon General, had served as chief surgeon of the AEF in France during World War I, until his appointment as Surgeon General in the fall of 1918. He realized the great importance of such studies, which he stressed frequently.
WAR CASUALTIES AND HOSPITAL REQUIREMENTS
It was thought possible to present statistically the experiences of the Medical Department during World War I, both in the Zone of Interior and in the theaters of operations, so that these experiences might be of assistance in planning the hospital and
86
medical service requirements for any future similar emergency. In addition, such a study could be particularly significant in determining the requirements for replacements both in the Zone of Interior and the theaters of operations.
In the period between the close of the Spanish-American War and the beginning of World War I, earnest, intelligent, and persistent study was made of this subject. War experience of sufficient magnitude was unavailable, however, upon which to base such a study.
Two excellent studies with reference to the sanitary service and hospital requirements were published in 1910 by Maj. Paul F. Straub1 and by Maj. John F. Morrison and Maj. Edward L. Munson.2 These were followed by a third study published by Major Munson in 1911.3
In his study, Major Straub described, carefully and with great skill, the principle of the collection of casualties and their evacuation to field hospitals, as used by Maj. Jonathan Letterman during the Civil War. Major Straub included evacuation hospitals and convalescent camps, in addition to ambulance trains, for the evacuation of the sick and wounded from the forward hospitals to those in the rear which had not been included in Major Letterman`s plan. During 1862-63, a lull generally followed each major battle which allowed sufficient time for the ambulances used to evacuate the sick or casualties from the collecting stations and field hospitals and to transport the patients to the hospitals in the rear. After the Battle of the Wilderness, Spottsylvania Courthouse, and others that followed, when the expected lull did not appear, Major Letterman`s ambulance trains were inadequate both to evacuate the casualties from the field hospitals to those in the rear and to remain in service near the frontlines. In addition, the Wilderness baffle area had no evacuation hospitals to which the patients could be moved with the result that field hospitals were not available promptly for use at the Spottsylvania Courthouse and other subsequent battle areas.
1Straub, Paul F.: Medical Service in Campaign. Philadelphia: P. Blackiston`s Sons & Co., 1910.
2Morrison, John F., and Munson, Edward L.: A Study in Troop Landing and Management of the Sanitary Service in War. Fort Leavenworth, Kans., 1910.
3Munson, Edward L.: The Principle of Sanitary Tactics, A Handbook on the Use of Medical Department Detachments and Organizations in Campaign. Menasha, Wisconsin: Banta Publishing Co., 1911.
87
Major Straub, in his monograph, included a mathematical formula to estimate the expected number of casualties in a hospital by days. This formula was developed by two German medical officers, Cron and Hochman, from experience with the German Army during the Franco-Prussian War of 1870-71. The formula, when expressed in the conventional way as an algebraic equation, was shown to be a third-order parabola. As expressed for the battle casualties, the number of patients increased quite rapidly in hospitals to reach a maximum number on the 50th day. After a few days of stabilization, the curve of expectancy dropped in a manner indicating a rapid decrease in the number of patients after that time. Such a conclusion may have been warranted by the actual hospital experience at that time. The Franco-Prussian War was of short duration, with a number of major military engagements in rapid succession and with heavy losses, except for the siege of Paris. The experience of the American Army during World War I was not similar to this situation. As a result, little use, if any, could be made of Major Straub`s German formula during World War I.
Consequently, little, if any, experience was available to the medical officers for an estimate of the number of hospital beds required in fixed hospitals. Although the General Staff officers were anxious to authorize a sufficient number of such beds, they knew that transportation was so urgently needed for many other things that they did not feel that it should be provided for unnecessary hospital equipment or personnel.
Although both the medical officers and the General Staff officers were equally conscientious and anxious to provide adequate hospitalization, it was exceedingly difficult to prepare a reasonable estimate from the data available. Fortunately, the experience of the English and French Armies during the several years in which they had been engaged in the conflict before the United States entered the war, was of the greatest assistance to both the American medical and General Staff officers in solving the problem.
In 1928, Col. A. P. Clark, MC, while on the General Staff, published a study4 which described a plan to assist in arriving
4Clark, A. P.: Replacements and Hospitalization and Their Relation to Losses in War. Mil. Surgeon 63:313, September 1928.
88
at a more accurate estimate of medical and hospital requirements. Unfortunately, his plan just was not feasible.
Love was a student in the School of Hygiene and Public Health of the Johns Hopkins University during the session in 1927-28. In the spring of 1928, he discussed with Dr. Lowell J. Reed,5 professor of biostatistics, the use of the World War I experience of the U. S. Army in the development of a satisfactory statistical study which might meet this need. He called Dr. Reed`s attention to tables in the annual report of the Surgeon General showing the duration of treatment of Army patients in Army hospitals during the years of 1925, 1926, and 1927. These tables showed duration treatment by 7-day periods. Dr. Reed suggested that this experience might be used by graduating the data to a suitable curve and fitting a mathematical equation to the curve. The equation for the curve could then be integrated to show the number of patients expected in hospitals from day to day.
The method of handling the data can be described briefly as follows:6
1. The combined hospital experience for the 3 years was consolidated into one total. This total experience was reduced to a per mil basis to show the number of patients in each 1,000 who received treatment for 1, 2, or more weeks.
2. These data for hospital patients by weeks were then converted to patients remaining in hospitals after each week by deducting the number leaving during that week from the number that had remained from the preceding week. After the 14th week, the number remaining was found to be small and progressively smaller thereafter. They were, therefore, placed into one group. Obviously, the number leaving the hospital each week included those who returned to duty, died, were discharged for disability or otherwise left the military service.
3. An attempt was then made to fit the material to a second-or third-order parabola and also to certain types of Pearsonian curves, but the results were not satisfactory. It soon became apparent that there was a very definite geometrical progressive
5Dr. Reed was later dean of the School of Hygiene and Public Health; then vice president of the Johns Hopkins University in charge of medical and medical educational facilities; and finally president of the Johns Hopkins University, from which office he retired in 1956.
6Love, A. G.: War Casualties, Their Relation to Medical Service and Replacements. Army M. Bull. No. 24, 1931.
89
relationship between the group of cases leaving sick report by time intervals to those remaining. This relationship determined that the curve was of the exponential type.
4. Inspection showed that this type of curve fitted the observed material quite well with only a few exceptions. As developed later, the exceptions were usually due to imperfections in or to incompleteness of the basic material. On the whole, the exponential curve not only best fitted the observed points but it graduated also satisfactorily the grouped experience beyond the 14th or the 20th week.
After it was determined that the exponential curve was suitable for graduating the material for the sick and injured patients who had been treated in the Army hospitals during 1925-27, it was necessary to reassemble the experience during World War I by some definite duration of treatment period. In these additional tabulations, 5 days` duration of treatment was used instead of 7 days. It was necessary to rerun the cards from the 1917-18 period to obtain the following information:
1. All cases wounded by gunshot missiles and all deaths and disability discharges resulting therefrom.
2. All cases wounded by poisonous gases and all deaths and disability discharges resulting therefrom.
3. (1) A sample of approximately 150,000 out of about 450,000 disease and nonbattle injury cases in the AEF during the last 6 months of 1918. The sample included the same proportion of each diagnosis insofar as practicable. (2) All the deaths from disease and nonbattle injury in the AEF during 1918.
4. All deaths and discharges for disability occurring among troops in the United States during 1918.
The task of tabulating a sufficiently large sample of the duty cases in the United States during 1918 was too great to be undertaken with the time and personnel available in 1928-29. However, as has been stated, similar data were available for the troops in the United States during the years 1925-27, and it was assured that there was no material difference in the duration of treatment for such cases and of similar ones during 1918. Consequently, the data, as assembled for the 1925-27 cases, were used as 1918 experience with some necessary, slight modifications.
Exponential curves were fitted to the material thus obtained,
90
as for the peacetime experience. In some instances where the material was homogeneous, as in cases of poisonous gases or gunshot missiles, a one-section curve was sufficient. In another class of cases, particularly hospital cases of disease and non-battle injury, certain cases, such as compound fractures of the femur or pulmonary tuberculosis, which required prolonged treatment, were not in the same proportion as those needing only a few days` treatment in hospitals. In such instances, when there were relatively more short- than long-duration curves, a two-section curve was required with the general formula. In a third class of cases, such as those including patients treated in hospital or in quarters in the United States in 1925-27, the short-duration treatment cases, those treated in quarters, were relatively so much more numerous than the medium- and long-duration cases treated in hospitals that a three-section curve was required.
Although parts of this discussion may appear to be unnecessary and highly technical in character, the basic philosophy is quite simple. Cases of disease and injury occur from day to day. Patients who are admitted to sick report remain under treatment for varying periods of time, depending chiefly upon the severity of illness or injury but also, to a certain extent, upon such circumstances as proximity of the hospital to the troop areas and facilities for returning men to organizations. Of those admitted to sick report, some will have been returned to duty by the end of the first period, others will leave during the second, and others will remain a longer time, a few even a year or more. Therefore, during the second period, in addition to the new cases admitted, a certain number of patients will be remaining from those admitted during the first period. Similarly, during the third period, in addition to new admissions, some will be remaining from the first and second ones, and so the increase in the number of patients will continue as time advances until all the first group of cases admitted have left sick report, which for all practical purposes will be at the end of 1 year. After that time, the number of patients leaving sick report during each period by return to duty, death, discharge for disability, or other disposition will equal the number of patients admitted.
When the basic equation was solved after the addition of suitable constants, the results showed the number of patients, including the remaining and current admissions. All the basic
91
curves showing the daily increase in the number of patients in hospital, both in the Zone of Interior and the theater of operations, were developed on the basis of the admission of 1 patient per day per 1,000 strength. This was done so that any strength or any admission rate that it was decided desirable to use might be used to multiply the data shown for any position on a curve.
After these basic curves were developed, others were added to show the daily number remaining in the hospitals from time to time for each type of case and also those who were returned to duty. In addition, tables were added to show the effect upon the number in hospital of the evacuation policy that might be decided upon as desirable for patients from the theater of operations to the Zone of Interior.
An attempt was also made to show the effect upon the flow of replacements back to the troop areas of patients from hospitals, as affected by any system of evacuation from the theater of operations to the Zone of Interior that might be decided upon. It was felt that such cases as could be returned to duty within a reasonable period of time should receive their treatment in the theater of operations rather than being evacuated from that area. This obviously was an attempt to demonstrate the necessity, or rather the economy, of providing an adequate number of beds in fixed hospitals in a theater of operations, since it was cheaper to transport medical material and personnel to the theater than to train and ship replacements to that theater.
After this was done, an attempt was made to show, by graphs and in tables, the experience of the U. S. Army for the period of the Civil War and the years following until World War I. World War I experience included the treatment of both ordinary casualties in the Zone of Interior as well as in the theater of operations. It included, in addition, battle losses by engagements and by different strengths of organization, such as regiment or divisions, corps, and army.
It was hoped that this experience might be of assistance to those who were attempting to estimate hospital requirements in any future military expansion similar to the one in World War I. It was, of course, recognized that the experience of World War I must necessarily be modified to meet the conditions that existed at the time the estimates were being made.
92
In a further study of the war casualty experience, three tables were constructed and published in the Army Medical Bulletin, No. 55, January 1941. They showed the following:
1. Number of officers and enlisted men (combined) wounded by gas and by gunshot missiles, with the total for the two, and the number killed in action, by component organizations of divisions and brigades and by engagements.
2. Number of officers and enlisted men (combined) wounded by gas and by gunshot missiles, with the total for the two, and the number killed in action, by divisions and by engagements.
3. Number of officers and enlisted men (combined) wounded by gas and gunshot missiles, with the total for the two, and the number killed in action, by engagements, by divisions, and by nondivisional units.
The totals in these tables differed somewhat from similar tables prepared by the Office of the Adjutant General and published in the Annual Report of the Secretary of War for 1926. The Adjutant General`s tables were prepared apparently from reports of the several organizations, whereas those compiled in the Surgeon General`s Office were from names of organizations and dates, as shown on the sick and wounded report cards of the individual patients. Obviously, there would be some difference in the data compiled in the two separate ways. It is hoped, however, that the tables published in the Army Medical Bulletin, No. 55 would be of service, in addition to those published in the Secretary of War`s report of 1926.
It must again be stated that this work was made possible by Surgeon General Ireland whose thorough knowledge of the military medical organization and sympathetic appreciation of the various problems that were connected with it were a constant help and inspiration to all members of the Medical Department. His assistance was given in every way possible.
Biometric Studies on U. S. Army Officers
The second important study that was considered desirable was one of the physical condition of Army officers when commissioned and during their service: their physical measurements and the relation of these observed facts to their health, efficiency, and life expectancy. It had been recognized for a
93
number of years that such a study was desirable. Civilian life insurance companies were hesitant to insure officers at normal rates, fearing the effects of tropical and other hazardous service in times of peace and service in time of war upon the life expectancy of the officers. In addition, information was desired with reference to certain aspects of the physical examination standards, such as height, weight, chest measurements, blood pressure, and so forth, the changes in them over a period of years, and their relation to health.
Army officers had careful physical examinations when commissioned and for many years thereafter at irregular intervals when promoted. Since 1907, each officer has had a careful physical examination at least once a year.
Quantitative data in regard to stature, weight, chest measurements, pulse rate, and systolic and diastolic blood pressure7 were recorded on the reports of these examinations, and notations were made in regard to the physical condition of the officers. In addition, the medical report cards showed any sickness or injury that occurred and any medical treatment. Death occurring after retirement was reported, with the cause, to The Adjutant General. There was thus an opportunity to observe these anthropometric measurements with any change in them and in their interrelation which occurred with time and, in addition, their relation to health and life expectancy.
During the latter part of the scholastic year 1927-28, Love discussed such a study with Dr. Lowell J. Reed, professor of biostatistics at the Johns Hopkins University, and with Dr. Halbert Dunn, his assistant. They agreed to participate jointly as representatives of the department of biostatistics with the Surgeon General`s Office, provided the Adjutant General`s Office would authorize the examinations and use of the necessary records. Unfortunately, for the purpose of the study, Dr. Dunn was unable to participate since he left the school to accept a position elsewhere before the work was started.
General Ireland, who always supported any project that might advance the interests of the medical profession and assist the Medical Department in improving the service to the Army, promptly approved the plan. With his approval, The Adjutant General was requested to grant authority for the use
7Blood pressure recordings were begun in 1916.
94
of the officer personnel files. He consented, and the personnel in his office gave constructive help throughout the undertaking.
An exhaustive examination and study were made of all records available. The records examined included those of the physical examinations when the officer was admitted to the U.S. Military Academy and when he was commissioned, the annual physical examinations, the physical examinations upon promotion or retirement, all records of hospital treatment, and the cause of death after retirement. All essential data were extracted upon suitable forms. This phase of the work required about 1? years.
It had been planned to include in the sample all officers who were in the active service on 1 January 1901, in addition to those who were commissioned between that time and 31 December 1916. Since this group was larger than could be handled with the available personnel, the names selected included all those whose surnames began with the letters A to P, inclusive. The total sample included 5,021 officers.
Practically all the officers whose records were studied were 21 years of age or over when commissioned. The physical examination and medical data were generally then for men above that age, except for the records of the entrance examination of those who had been commissioned from the U.S. Military Academy. The actual number of years that any officer was under observation by these records was 28? years; that is, from 1 January 1901 to about June 1929 when the abstraction of the material from the files was concluded. The average time the officers were on the active list was 17? years, and the average time, including that on the retired list, was 20 years. Since, however, the data were included from physical examinations recorded many years before 1901, even as early as in the 1870`s, the actual period of observation was much greater.
As a result of this study, Reed and Love published four papers under the general title `Biometric Study on United States Army Officers.`8 These papers were issued jointly from the Department of Biostatistics of the School of Hygiene and Public
8(1) Reed, L. J., and Love, A. G.; Longevity in Relation to Physical Fitness. Proceedings of the 21st annual meeting of the medical section, American Life Convention, Washington, D. C., 1931. (2) - -: Somatological Norms, Correlations, and Changes with Age. Human Biol. 4(4): 509-524, December 1932. (3) - -: Somatological Norms in Disease. Human Biol. 5(1): 61-93, February 1933. (4) - -:
Economic Efficiency (Length of Service) in Relation to Physical Fitness and Other Factors. Mil. Surgeon 71: 231-238, September 1932.
95
Health of the Johns Hopkins University and the Office of the Surgeon General. A brief review of the findings may be of interest to those who are not conversant with the studies.
Longevity in relation to physical fitness.-The officers, some of whom were observed by means of records from 1 January 1901 to 30 June 1929, with a maximum period of 28.5 years, an average one of active service of 17.5, and an average one of life after retirement, including those on the retired list, of 20 years, had an average expectancy of life greater by 2 years than a civilian at ages 21 to 61, inclusive; 1.5 at 71 years; and 1 at 81 years. If the selection by physical examination affects the expectancy of life materially only during the first 5 years, as it is understood that life insurance examiners claim, and after 10 years only slightly if at all, it is apparent that there must be some other factor, besides the initial physical selection, to account for the greater longevity of officers compared with civilians which is evident even up to age 81. This increased expectancy of life is especially noteworthy in view of the extra hazard to which officers were subjected while serving in the Tropics, especially so in the Philippine Islands during the early years of this experience and also in view of the additional hazard of military combat. Possibly it may be said that repeated physical examinations, with resulting prompt initiation of indicated remedial measures and the routine physical exercise in which officers engage, actually do so materially improve the health of the officers as to increase longevity.
Correlations and changes with age.-As has been noted by a number of observers, men do not attain their full growth until the age of 25 or even later. Thus, the difference between the means of the stature or height at ages 21-25 and 26-35, which was 0.22 inches, although small, is apparently a significant one. After the ages 26-35, there was a slight, though probably not a real, decrease in stature as measured in inches.
Weight, and especially weight in relation to height and age, is an important biometric measure. The average weight of the Army officers considered with a stature of 70 inches was 151.27 pounds at ages 21-25, which increases gradually through various age groups to reach a maximum of 169.81 at ages 56-64. There is frequently an increase in weight after the age of 30, but, as has been pointed out by life insurance examiners and actuarial offices, the average ideal for weight at age 30 is the
96
desirable one and one that should be maintained practically throughout life. The increase in weight above the ideal average, after 30, is both unnecessary and undesirable.
It was interesting to observe that officers who were markedly above the average for age and height at ages 46-55 were also above the average at ages 21-25. Thus, the heavy group of officers, who weighed 151.86 (average 146.66) pounds at 21-25, weighed 186.27 (average 164.02) in the age groups 46-55; whereas the light groups who weighed 132.16 in the age group 21-25, weighed only 132.49 pounds in the age group 46-55.
Blood pressure, as has been stated, was not recorded in the annual physical examinations with any regularity until 1916. From such data as were available, however, it was observed that there was an increase in the blood pressure as age advanced. Thus, with the stature or height of 70 inches in the age group of 21-25, the systolic was 122.28, and in the 56-64 age group it was 139.37. There was also an increase in the average of the diastolic blood pressure; thus, with a stature of 70 inches at age 21-25 it was 81.44, and at 56-64, 85.82.
The first study showed the following conclusions:
1. After 25 years of age, no significant change occurs in the mean either of the stature or the pulse rate.
2. A significant, but progressively declining, increase in the mean weight and chest measurements was apparent. This increase, which is generally distributed, occurs almost entirely in the late 20`s, 30`s, and 40`s, and there is little, if any, change after age 55.
3. A significant increase in the mean and median of both the systolic and diastolic blood pressures was noted. The change, however, is not material until late in life and becomes apparent chiefly in the 50`s and 60`s.
Somatological norms in disease.-The relation of certain physical measurements to the occurrence of disease was described in the third paper. During a discussion with Dr. Reed concerning the probability value of the interrelated height, weight, and chest measurement (exp.) of selective service men, found at camps, to pulmonary tuberculosis or valvular heart disease of early life, Dr. Reed suggested the use of the formula of the normal correlation surface to graduated height-weight, and height-chest (exp.) frequencies in the general group and also in each of the two diseased groups. From the data thus graduated, the
97
percentage distribution of each diseased group in the various height-weight and height-chest groups could be calculated.
The results of the study9 showed that the two diseased groups occurred more frequently among thin men and, particularly, among tall, thin men. Thus, the men who were 74 inches in height and 30 pounds below the average weight for their height showed twice as many cases of valvular heart disease as did those of average weight for height. However, for those who were only 70 inches tall and 30 pounds below average weight for height, the probability was increased only 1? times, and those 66 inches in stature and 30 pounds below average weight for height showed only one-third more cases.
For the men with pulmonary tuberculosis, the picture was even more striking. Among men who were 74 inches in height and 30 pounds below average weight for height, there were 10 times as many cases of tuberculosis as in the general group; for those 70 inches in stature and 30 pounds underweight for height, there were 6 times as many cases of tuberculosis; and for those 66 inches tall and 30 pounds underweight for height, there were only 3 times as many cases.
The formula for the normal correlation surface was applied also in the study of the relation of the biometric measurement of officers in relation to disease groups. As among Selective Service registrants, so among officers, pulmonary tuberculosis and valvular (rheumatic) heart conditions occurred more frequently among tall, thin men. Thus, the weight of those at ages 21-25, who developed pulmonary tuberculosis 20 years later, was 140.25 pounds, with a stature of 70 inches, as compared with the average of 151.27 pounds in the general group. The difference of 10.88 pounds is certainly a significant one. Similarly, at ages 26-35, the weight of those who developed the disease 10 years later was 139.73 pounds, with a stature of 70 inches, as compared with 157.21 pounds in the general group, a difference of 18.48 pounds. It was found that officers who ultimately developed pulmonary tuberculosis and who had the characteristic tall, spare build had the same build from 7? to 20 years before the disease became clinically evident. During this experience of 5,021 Army officers with 100,000
9Love, A. G.: Somatological Norms in Tuberculosis and Heart Disease. Human Biol. 1: 166-197, May 1929.
98
years of life exposure, pulmonary tuberculosis resulted in 65 retirements from the active service, 13 deaths while on active duty, and 34 deaths after retirement.
Similar to those with pulmonary tuberculosis, the officers who showed evidence of valvular diseases of the heart which occured as the result of acute rheumatic fever and other infectious diseases were tall and spare during the first 20-25 years of life. This group, like the tubercular one, is of special interest on account of the persistence of the build of the patient. The mean weight and stature of the officer patients with this disease was about the same as that of Selective Service men of the same age who showed the disease. Thus, at ages 21-25, the officers or whom such a clinical diagnosis was made either then or later in life had a weight of 145.84 pounds, with a stature of 70 inches, as compared with 144.79 pounds for the Selective Service men between the ages of 21 and 30 who had the disease when they went to camp. In addition, the mean statures of officers with valvular diseases of the heart was uniformly greater than that of the general group. Not only were such patients tall and thin throughout life, but the distribution of the weight as measured by the standard deviation was less than that of the general group, indicating, as in the case of pulmonary tuberculosis, the compactness of the group. Valvular diseases of the heart, however, were a minor cause of disability and death during this experience. Forty-one of the officers were retired as disabled and seven died while on the active list, making a total of 48 separations from the service. In addition, nine died as a result of this cause after their retirement.
The diseases that were classified as the degenerative cardiovascular diseases, namely, myocarditis, arterial sclerosis, arterial hypertension, nephritis, apoplexy, coronary sclerosis, coronary thrombosis, and angina pectoris, during this experience, were of very great importance as a cause of death and disability.
The degenerative cardiovascular renal diseases, combined with the valvular heart diseases that occurred as the result of infection during early life, were the cause of 396 deaths during this experience or 36.8 percent of the total number which occurred. The ages 21-90 death rate among officers as a result of this group of diseases was 3.48 per 1,000 years of life, as compared with 3.65 among the white males of the same age group
99
in the U.S. registration area during 1919-21.10 This group of diseases was also responsible for 218 or 24.7 percent of the 881 retirements for disability before age 64, the retirement rate being 2.49 per 1,000 years of active service, 2.32 for nervous and mental diseases, and 0.85 for tuberculosis of all forms. The serious character of the cardiovascular renal diseases is shown by the fact that after retirement the expectancy of life at every age among those affected was only half as great as that of the general group.
The degenerative cardiovascular renal diseases, excluding the infectious valvular diseases of early life, caused the death or disability retirement of five times as many men as did military combat during the period covered by the experience, which included the last 1? years of the Philippine Insurrection, subsequent guerilla warfare in that country, and World War I. They were responsible for the retirement or death on the active list of 254 officers, or 19 percent of those from all causes. Including the deaths which occurred after retirement, they caused 35 percent of the total and were responsible for twice as many as cancer, tuberculosis, and military combat combined.
It was, therefore, apparent that the group of cardiovascular renal diseases was one that warranted very careful study for those factors that might be of assistance in arriving at an early diagnosis in order that corrective measures might be instituted. Consequently, special attention was devoted to the study of the height, weight, and other factors of officers in relation to the expectancy of the occurrence of these diseases.
In the paper, Somatological Norms, Correlations, and Changes With Age, there was a discussion of the material increase in the mean weight with the advance in age. It was found that the greatest increase in weight occurred during the 30- and 40-age period with but little, if any, after age 55. At age 21 to 25, there was no material difference between the mean weight of the general group and those who developed the cardiovascular renal disease at that age. Similarly, at age 26 to 35, no real difference was found between the weight of the general group and that of those who developed the disease at that age; but there was a significant increase in weight among those
10See footnote 8(1), p 94.
100
who developed the disease 10 to 20 years later. Beginning with the age period 36 to 45, there was a real difference, at every age, between the average weight of the general group and that of the cases diagnosed either then or 10 to 20 years earlier. Thus, the officers who showed the first clinical evidence of this degenerative type of disease during the decade from age 56 to 64, weighed at that time 177.47 pounds, with a stature of 70 inches; 10 years earlier, they had weighed 178.91; and 20 years earlier, 177.46, with the same stature. These weights compared with weights of 169.81, 168.19, and 164.10 pounds for the general group. Similarly, the ones who died between the ages of 66 and 75 as the result of one of these diseases, had at ages 56 to 64 a weight with a stature of 70 inches of 181.05 pounds and at ages 46 to 55 a weight of 183.64 pounds, as compared with 169.81 and 168.19 pounds, respectively, for the general group. In addition, the mean stature of the officers with such a disease was uniformly lower than the one of the general group. Although, in the majority of instances, it is not significantly different when tested with its probable error, the persistence of the difference apparently indicates a real one.
Of the officers in the age group 46-55, 11.47 percent, either at that age or later in life, developed a cardiovascular degenerative disease. In a study of the distribution of the disease in the height and weight zones, it was found that with a stature of 70 inches, 7.5 percent of those at 156 pounds had, either during that age or later in life, a cardiovascular-renal disease, as compared with 75 percent of those weighing 230 pounds. Officers with an average weight for age had the disease expectancy of about 9 percent, as compared with 8 percent for those who by diet and exercise had retained the desirable weight-stature; that is, the weight of the 26-35 age group. On the other hand, an increase of 15 pounds above the average for the age raised the disease expectancy to 13 percent.
Since the systolic blood pressure was first recorded consistently on the reports of the physical examinations in 1916, the maximum period of observation of this variable in any group in this study was 14 years. The results showed that little increase in the mean of the systolic blood pressure was observed in this experience until after 45 years of age and that 50 percent of the total increase occurred between ages 56 and 64. Further, it was shown that when the known diseased cases were
101
included, the means for the 21-25 to 56-64 age groups were 122.04, 123.16, 125.29, 129.15, and 136.58, respectively. The percentage of the diseased cases in the general group increased from 1.7 at age 26-35 to 4.1 in the next age group, then to 8.9, and finally to 16.4 in the 56-64 age group. At 26-35, the systolic reading for officers who had a cardiovascular renal disease, either at that age or later in life, was 9.55 mm. greater than that of the general group. The difference increased to 14.15 mm. during the next age group, then to 17.72, after which there was a slight decline to 16.71 at ages 56-64. If, however, the mean of the diseased group was compared with that of the general group, after excluding the known pathological cases, the difference in the readings was 9.71, 14.77, 19.47, and 19.90, respectively.
As with the cardiovascular diseases of degenerative character, it was found that there was a marked increase in the systolic blood pressure with the increase of weight for stature, in the age groups studied. The cardiovascular group had a characteristic increase in weight as much as 20 years before the disease became clinically apparent. In the cases with increased systolic blood pressure, however, covering only 14 years (which was the maximum time of any one case), the increase in the systolic blood pressure did not become apparent until 10 years, or even less, before the clinical diagnosis was made.
As with the systolic blood pressure, there was found but little increase in the diastolic blood pressure of the general age group until after age 45, and more than 50 percent of the total rise occurred after age 55. When all the known pathological cardiovascular renal cases were excluded, the means by age groups from 26-35 to 56-64 were 80.33, 80.52, 82.21, and 84.60 mm., respectively. At each age, the diastolic blood pressure of the diseased group was materially greater than that of the general one, the difference being, beginning with age group 26-35, 5.38, 11.41, 10.25, and 8.05 mm., respectively.
A very significant rise was found in the mean of the diastolic blood pressure 10 years before the disease was apparent clinically, but, like the systolic, the diastolic continued to increase until the age when the disease was diagnosed. Similar to that observed with the systolic blood pressure, there was a material increase in the diastolic blood pressure as the weight by stature increased in the several age groups.
102
During the experience with the 5,021 Army officers, cancer and other malignant tumors caused 84 deaths. So far as could be determined from the limited number of cases under observation for which height and weight data were available (68 cases), it was found that the stature of cancer cases is less than that of the general group. On the other hand, so far as could be determined, there was no uniform difference between the weight-stature or height relationship of the cancer cases, as compared with that of the general group.
Nervous and mental diseases were the most frequent causes of permanent disability during the experience. These diseases caused a total of 215 retirements for disability, as compared with 140 for diseases of the circulatory system. When, however, the nephritis cases were added to the diseases of the circulatory system, the total for the two groups was 218, or 3 greater than the retirement for nervous and mental diseases. Of the 215 cases retired for nervous and mental diseases, 133 were for neurasthenia, psychasthenia, and psychoneurosis as then diagnosed; 46 for psychosis; and the remaining for scattered causes. Experience showed that officers who were retired as a result of such conditions had, at every age, an expectancy of life as great as that of the general group. No significant or uniform difference was found between the mean stature of the officers in the mental and nervous group and that of the general group. There was, however, a persistent, uniform, and, apparently, a real decrease at every age in the mean weight of the men who developed such a disease either then or later in life.
The conclusions of the study on the Somatological Norms in Disease were as follows:
1. Cases of pulmonary tuberculosis in young men were found to be much more frequent among thin men, and especially tall, thin men.
2. Only about 1 percent of the officers developed an active pulmonary tuberculosis after age 46, but the majority of those who did were tall and spare.
3. Tuberculous individuals had a characteristic build which could be noted as much as 20 years before the clinical evidence of the disease appeared.
4. Valvular diseases of the heart, which result from acute rheumatic fever and other infectious diseases early in life, were found to be more frequent among tall, thin individuals at advanced
103
ages as well as in early adult life. The deficiency of weight in such cases was, however, not so pronounced as in those with tuberculosis.
5. Cardiovascular renal diseases of the degenerative type, which develop in later life, were the most frequent causes of disability and death. Of the officers who were active at age 46, 11.4 percent subsequently had such a disease so diagnosed.
6. Those individuals who were overweight after 30 years of age had a much greater disease expectancy than those of average or less than average weight.
7. The increase in weight was apparent from 10 to 20 years before the disease had fully developed, with no additional increase after the disease was diagnosed.
8. The excess in weight above the average for age 30 was found to be a most significant factor.
9. Since both underweight and overweight are undesirable, the ability to maintain the age 30 average weight can be said to increase the probability of a maximum of health and longevity.
10. Cardiovascular renal diseases were associated with an increase in the systolic and diastolic blood pressure. The disease expectancy increased rapidly as the blood pressure became greater.
11. In this experience, some of the increase in the blood pressure was apparent 10 years before the disease was diagnosed but continued to increase with the development of the disease.
Economic efficiency (length of service) in relation to physical fitness and other features.-In the study of physical fitness in relation to longevity, some interesting facts were discovered. During this experience with an average of 17.5 years on the active list, 2,900 of the 5,021 officers were separated from the service for various causes. Resignations, which were the most frequent cause during the early ages, were responsible for 15.1 percent of the total separations. Disability retirements were responsible for 30.4 percent; death, 16.0 percent; court-martial, 3.1 percent; class B separations, 2.3 percent; separation by the act of 1922, 4.3 percent; desertions, 0.4 percent; retirement after 30 years` service, 10.2 percent; retirement after 40 years` service, 3.6 percent; retirement at age 64, 13.1 percent; and other causes, 1.6 percent. The total for length of service and age was 26.9 percent.
104
The separation rate per 1,000 man-years of service was 33.18. Almost one-third of the total was caused by disability retirements, with a rate of 10.08 per 1,000. The deaths while on the active list stood second, with 5.30 per 1,000; and resignations third, with 5.00 per 1,000. The combined rate for separation due to retirement for length of service and age was 8.92 per 1,000, which was exceeded only by the rate due to disability retirement.
Officers who entered the Army at age 21 had an average expectancy of service of 26.6 years, with a possible maximum of 43 years. Similarly, at age 44, they had an average expectancy of 13.1, with a possible maximum of 20 years.
The conclusions as drawn from this study on length of service in relation to physical fitness were as follows:
1. The average age of Army officers when commissioned was approximately 25 years.
2. During this experience with 5,021 officers, with an average period of observation on the active list of 17.5 years, 2,900 separations occurred. Of the 517 separations between ages 61 and 64, a total of 465 was due to age and length of service.
3. Retirements for physical disability and death caused 1,293 of the 2,383 separations between ages 21 and 60. Physical fitness was, therefore, not only closely related to longevity but also to economic efficiency.
4. Degenerative cardiovascular renal diseases were responsible for 167 disability retirements and 90 deaths, a total of 257 or 19.1 percent of all retirements for physical disability. Since these diseases were observed to occur much more frequently in the overweight class, a portion of the resultant economic waste might be prevented if individuals accepted the fact that the average weight at 30 years of age is the most desirable weight to maintain thereafter and that excess weight results in physical impairment and shortening of life.
5. During this period, which included 1.5 years of the Philippine Insurrection, the subsequent guerilla warfare, and World War I, 31 officers lost their lives as the result of injuries in military combat and 22 were permanently disabled, making a total of 53 separations due to military combat. Cardiovascular renal diseases can, therefore, be said to have killed or disabled almost five times as many men as military combat.
105
6. Similarly, underweight among the younger officers and in some instances among the older ones was found to be associated with a shortening of life and a lessening of economic efficiency. Pulmonary tuberculosis, which occurred much more frequently among those undernourished, caused 65 retirements and 13 deaths. Some of this loss might be preventable by more efficient attention to the nutrition of the individuals.
7. Mental and nervous diseases were responsible for 23 percent of the disability retirements, resulting in 20 percent among all officers in the Army during the last 10 years.
8. Diseases of the ear, with defective hearing, resulted in the retirement of 58 officers, only 7 less than from pulmonary tuberculosis. On the other hand, eye conditions were responsible for only 18 retirements. Defective hearing alone caused 48, as compared with 3 from refractive errors.
ANTHROPOMERIC MEASUREMENTS IN RELATION TO HEALTH
The third study undertaken was the preparation of tables of weight and chest measurements in relation to height by age groups for the guidance of medical officers in the examination of recruits, U.S. Military Academy cadets, and officers. The study was started in 1926 by Love and continued from that time until 1932.11
The first height-weight chest measurement standards for recruits in the Army were published in 1887 and were in force from that time, practically unchanged. Similar standards for admission to the U.S. Military Academy were adopted in 1894 and revised in 1904. They were still in use in December 1921. The standards for weight and chest measurement in relation to height for ages of officers, until December 1931, were based upon the findings of the medical actuarial investigation of 1912.
The experience with the Selective Service men at ages 21 to 30 during World War I was used for the study of the average weight and the spread of weight and chest measurements in relation to height among men of that age group (p. 82). The
11Love, A. G.: Physical Measurements-Their Relation to Health. Army M. Bull. No. 28, July 1932.
106
studies of Reed and Love, Biometric Studies on U.S. Army Officers, furnished similar material for the officer group.
To obtain similar data for the younger men, ages 17 to 21, inclusive, the records of physical measurements were studied for 100,000 young men of that age group. Approximately two-thirds of these men went to the Citizens Military Training Camps in 1925-26, and the remaining one-third were from the Reserve Officers Training Corps in 1921-22. Ample material was thus available for the setting up of the standards of the various age groups, from 17 for the young men to 64 for officers at the age of retirement. The constants of the height, weight and chest measurements, and their interrelation, were calculated and published in a table in the Army Medical Bulletin, No. 28, referred to previously.
The careful use of standards during physical examinations are manifold. Physicians and health agencies recognize their importance as a basis for preserving the health of the individual. Life insurance companies know that they are economically necessary to prevent excessive mortality risks among their insured; all organizations requiring arduous physical effort from their employees realize that only by the use of physical examinations in relation to standards can the unfit be excluded.
With reference to the selection of military personnel, Baxter12 quotes Vegetius as saying in his Art of War, `An Army raised without due regard to the choice of recruits was never yet made a good army by any length of service.` Colonel Baxter adds that it is indisputable that the extreme care in the selection of the material was one of the causes of the early invincibility of the Roman armies.
Satisfactory physical examinations should include, in addition to careful consideration of all physical signs and thorough use of refined diagnostic means, estimation of the physique based upon the weight and chest girth in relation to height. In examination of an individual, the finding of abnormal weight is a danger signal. With reference to its value to life insurance examiners, the late Dr. Brandreth Symonds, chief medical director of the Mutual Life Insurance Company of New York, said: "Only of late years have we become fully alive to the fact that the physique of an individual is a fundamental
12Baxter, Jedediah H.: Statistics, Medical and Anthropological of the Provost Marshal General`s Bureau. Washington: Government Printing Office, 1875, vol. I.
107
element in his selection as a life insurance risk. If a proper relation of weight to height and age is not secured when selecting a given group of risks, the mortality of that group will be high in spite of the utmost care in excluding all other unfavorable elements.`
Apparently the first table of weight for height was published in 1836 by Quetelet.13 One average weight and one average height was given for ages 20, 30, 40, 50, and 60. Such tables were gradually enlarged and additional details added. Among those published was one by Brent in 1884, one by Hutchinson in 1846, and another by Macaulay in 1881.
Hunter,14 of the New York Life Insurance Company, says that an American life insurance weight for height table was published in 1869; but, apparently, the first one based upon a considerable volume of statistics which took age into account was compiled in 1897 by Dr. George R. Shepherd. Several others based upon greater experience have appeared since that date. The reports of the careful and extensive Medico-Actuarial Mortality Investigation made in 1912 and published in New York in 1913, regarding those who had been insured between 1885 and 1908 by 40 life insurance companies in this country, was used quite generally as the basis for such tables in 1931. Colonel Baxter, who had been in the Office of the Provost Marshal General during the Civil War, was no doubt responsible for or influential in the publication of the standard table for weight and chest measurements in relation to height for recruits in the U.S. Army in 1887. With reference to this he said:
While the annals of recruiting contain copious detail as to stature, the amount of information furnished upon the subject of weight is for the most part extremely meager. A principal reason for this is to be found in the fact that weight is not a regulated quality in any code of laws governing the enlistment of recruits. The circumference of chest thought to be indispensable as an accompaniment to certain degrees of stature is carefully laid down in the English regulations, but weight is not even mentioned. It is to be presumed that the matter is left to the discretion of the examining surgeon, with whom the decision as to the other qualities named might, it is thought, be also left with advantage. A due proportion in the weight is quite as essential in the soldier as a well-formed
13Symonds, B.: The Mortality of Overweight and Underweight. McClure`s Magazine, New York, 1908-9, vol. XXXII.
14Hunter, Arthur: Journal of the Institute of Actuaries. London, 1912, vol. XLVI.
108
chest and is of greater importance than lofty stature. In former times, when it was necessary to make use of a ramrod in loading a musket, men of certain height were absolutely necessary for the service; but in these days of breechloading arms, a man from 5 feet to 5 feet 4 inches in stature and well proportioned in build and weight is, caeteris paribus, as serviceable a soldier as can be desired.
Reference has been made in the preceding discussion to the more frequent occurrence of pulmonary tuberculosis and valvular diseases of the heart from infections acquired early in life, among the thin men of the various ages even to an advanced age, and especially among the tall, thin men; but, in this connection, some additional experience may be of interest. Huntington said in 1876, "Forty deaths of underweights included no less than 25 of consumption and 5 of dropsy from diseases of the heart and liver. Consumption very, very often does send out its warning voice far in advance of its active coming, and in many cases this interference with the tissue making powers of the human system foreshadows the coming of tuberculosis.` Hunter,15 actuary of the New York Life Insurance Company, said, `At the age of entry 30-44, the death rate from tuberculosis was nine times as heavy among the underweights as among the overweights.` And Symonds adds, "The association of underweight and a tuberculosis family history has long been recognized as serious, especially in the younger ages. Thus we find that this combination gives a mortality of 180 percent in the ages below 35."
Insurance officials, in speaking of underweight and overweight, refer to those conditions as they existed at the time of acceptance of the insured as a risk. The report of the Medico-Actuarial Mortality Investigation in 1912 states:
The increasing ratio of mortality with decreasing weight at the younger ages of entry is noteworthy. The foregoing synopsis shows that for ages at entry 20 to 34, the insured who are from 5 to 10 pounds more than the average weight are better lives than those 5 to 10 pounds less the average weight * * *. At the younger ages, tall men have proved less desirable risks than short men. At the older ages the short and medium sized men have been slightly worse risks than tall men. As underweight has long been regarded as of serious importance at the younger ages and the presumption is that at these ages underweights have been selected with special care, emphasis is given to the somewhat unsatisfactory results disclosed.
15Hunter, Arthur: Journal of the Institute of Actuaries. London, 1913, vol. XLVII.
109
In the studies of Reed and Love referred to, it was found that even among officers of more advanced ages tuberculosis did occur more frequently among the thin officers and especially among the tall, thin ones. However, the diseases which were frequent among the older men, and especially the overweight ones, were those of the degenerative cardiovascular-renal group. That group was the most important cause of physical impairment and of deaths. The effects occurred principally after the age of 35, the incidence increasing as the men grew older. Repeating again for emphasis, in that experience,16 of the disability retirements and deaths among active officers, these degenerative diseases were responsible for the death or disability retirement of five times as many men as were injuries received in military combat during the period under study. If the deaths which occurred after retirement were included, these diseases caused 35 percent of the total and were responsible for twice as many deaths as cancer, tuberculosis, and military combat combined.
In a practical sense, it was desired to draw up tables, based on the results of the studies which would show a desirable weight-chest measurement for the various heights in the various age groups. After the basic data were compiled, showing the average weight and chest measurements for each height in the various age groupings, the next question to be answered was, "Is the average weight for each height the best standard or measure of health and vigor; and how far can one deviate from an apparently satisfactory one so as not to exclude too large a number of those whose physique differs from the average but who are healthy nevertheless, and at the same time not to include too many who are physically unfit or liable to early development of disease?` It was known that men who are healthy vary materially in weight for height, according to body conformation, bony framework, and general build. Hence, it was realized that permissible variation above and below the mean or the average weight for height should be established, which would make allowance for variation among healthy individuals and would, at the same time, furnish assurance of a reasonable degree of health and vigor. This, then, became a question of judgment, and the minimum and maximum
16See footnote 8(3), p. 94.
110
adopted for the selection of men for the military service would depend largely on how badly men are needed and what percentage of pulmonary tuberculosis or cardiovascular diseases of the heart the authorities feel that the situation warrants their assuming the risk of accepting.
In conclusion, the studies of Reed and Love added material experience to that accumulated by the life insurance companies. They added, also, additional data to establish probability zones. These studies, with the Selective Service experience and with that of the younger age groups as described in Army Medical Bulletin No. 28, furnish data upon which to base further realistic and meaningful standards. They demonstrate, in general, the great possibilities and flexibility of the punch card system with the use of the electrical accounting equipment and, in particular, the use made of the system by the Army Medical Department at a relatively early period.
