Vietnam, PTSD, Vietnam veterans, combat experience, military service, combat experiences, Vietnam veteran, Vietnam era, indicators, social problems, sleep disturbance, background characteristics, American Psychological Association, American Psychiatric Association, veteran group, veteran groups, national study, set of weights, defining characteristics, marital status, Frequent exposure, case weights, intrusive memories, Government Printing Office, psychological adjustment, American Sociological Association, Vietnam combat veterans, Residual stress, American Psychiatric Press, group differences, Vietnam era veterans, Vietnam combat
EPIDEMIOLOGY OF PTSD IN A NATIONAL COHORT OF VIETNAM VETERANS JOSEFINA J. CARD Sociomeirics Corporation Palo Alto, California At age 36, Vietnam veterans in the High School
class of 1963 reported significantly more problems related to nightmares, loss of control over behavior, emotional numbing, withdrawal from the external environment, hypcralertness, anxiety, and depression than did their classmates matched with them on 51 high school characteristics. These problems correspond closely to the disorder labeled post-traumatic stress disorder (PTSD) by the American Psychiatric Association. PTSD was associated with other family, mental health
, and social interaction problems. Some environmental variables--e.g., the presence of a spouse or being a churchgoer--were associated with reduced levels of PTSD or with reductions in the degree of association between combat and PTSD. The direction of cause and effect in these associations cannot be ascertained from our data, but it seems plausible to postulate that support factors can and do help some Vietnam veterans with PTSD. The study I am about to describe is based on the experiences of a group of men who were in the ninth grade in 1960. Most men in this class completed high school in 1963, I year before the official start of the Vietnam era. During the 1960s, one-half of these men served in the military, and one-fifth were sent to the Vietnam War
zone. This paper will focus on the antecedents and correlates of post-traumatic stress disorder among men in the class. A more complete description of the impact of the war on other aspects of the men's lives can be found in the book Lives after Vietnam (Card, 1983). The findings to be described are important because of their generalizability to a national cohort (age group) of men: The 1.1 million men who were in the ninth grade in I960 and who were still alive 21 years later, at age 36. While many other cohorts served in the decade known as the Vietnam era, the class of 1963 is an excellent one to study. The men in this class are very close to (within a year of) the average age of the Vietnam-era soldier. They graduated from high school a couple of years before the draft gained momentum. They contain large subgroups of both enlisted and drafted soldiers. A large proportion of the men were in Vietnam during the height of the fighting (the 1968 Tet offensive); thus, a wide range of combat experiences -- from none to heavy --are represented. Further, we will see how our study was a prospective one, which encompassed information on what the men were like prior to the military service of some, as well as information on how they developed after the transition back to civilian life. The availability of early prcservice information enabled us to determine, more accurately than was previously possible, the extent to which post-service differences between veterans and non-veterans were indeed attributable to the veterans' military or combat experience, as opposed to pre-existing differences between those who served and those who did not. No other national study of the Vietnam-era veteran has had this prospective, longitudinal feature. This research was supported by grant MH34643 from the National Institute of Mental Health. The author wishes to thank Drs. J. B. Jayme and W. S. Farrell for their helpful comments on an earlier draft of this paper. 6
Epidemiology of PTSD
METHOD Overview Three groups of men were surveyed: (1) the Vietnam veterans --a sample of approximately 500 men who served in the Vietnam war zone during the 1960s; (2) the nonVietnam veterans--approximately 500 classmates who served in the military during the same period, but who never were assigned to Vietnam; (3) the nonveterans -- approximately 500 classmates who never served in the military. All 1,500 men first had been studied as ninth graders (in 1960) and then again II years after high school (in 1974) as part of a larger effort known as Project TALENT. In early 1981, 6 years after the end of the Vietnam era and more than a decade after the veterans' military service, when the men in the class were approximately 36 years of age, we surveyed- them again for purposes of the present study. Variables The 1960 survey obtained comprehensive information on respondents' academic abilities, their vocational and avocational interests, their personalities, their home life, and their plans with regard to further schooling, work, and military service. The 1974 follow-up questionnaire included items on educational experiences and plans, careerrelated experiences and plans, military service, as well as marriage and family life, and health and community activities. The 1981 survey questionnaire developed for the present study included items that measured the following variables: Demographic background, marriages, children, educational history, job history,financialsituation, physical health, life satisfaction, alienation, social problems, anxiety, depression, hostility, brushes with the law, personality traits, organizational membership, reason for serving or not serving in the military, military experiences, and Vietnam experiences. The questionnaire was developed after a review of the literature and included a wide range of constructs that had been linked, either theoretically or empirically, to military service. Before the questionnaire was sent to members of the study sample, it was reviewed by a panel of experts on military service (a VA researcher, a Vietnam veteran, and a survey research expert) and was pretested on volunteer TALENT respondents who had not been chosen for the study sample of the present study. Sampling and data collection
Procedure The following steps were taken in selecting the participants in the current study and in collecting data from selected participants: 1. The men in the TALENT ninth-grade cohort were divided into two groups: (a) those who in 1974 responded that they had served or currently were serving on active military duty in the Army, Navy, Air Force, Marine Corps, or Coast Guard; and (b) those who in 1974 reported that they never had served on active military duty. 2. A random sample of 1,385 veterans, representative of all veterans in the class, were selected as the study's target-veteran sample. 3. A letter was sent to the 1,385 veterans in the target-veteran sample that explained the goals of this study, asked veterans to update the address Project TALENT had on file, and asked them whether they ever had served in Vietnam. (The existing TALENT file had information on whether the men ever had been in the military, but not on where such service was spent.) 4. Two follow-up letters were sent to those who failed to respond to thefirstletter. 5. Intensive attempts were made to locate those who failed to respond to the three mailings. These efforts consisted of consulting telephone directories and information operators, writing to postmasters, checking Department of Motor Vehicles records, and calling parents, relatives, and former neighbors. The effort was quite successful.
8Journal of Clinical Psychology
, January 1987, Vol. 43, No. 1
Of the 1,385 veterans in the target-veteran sample, 1,243 (90%) were located, 1,119 (81%) of whom had information on whether they had served in Vietnam. Seven veterans, located through their parents, were deceased; of necessity, they were dropped from the study. 6. The 266 (19%) men in the original target-veteran sample who could not be located or for whom information on Vietnam service could not be obtained and the 7 deceased veterans were replaced with other veterans. The replacements were chosen so as to match the originally chosen 266 nonrespondents on more than SO key demographic, cognitive, and sociopsychological characteristics. 7. There were 599 Vietnam veterans and 786 non-Vietnam veterans in the final target-veteran sample (originals plus replacements). All 599 Vietnam veterans and a subset of 566 of the non-Vietnam veterans were chosen as potential study respondents. 8. A group of 830 nonveterans then were chosen from the same TALENT ninthgrade cohort to serve as the comparison group. 9. Intensive attempts were made to locate the 830 members of the target nonveteran sample. These attempts were similar in nature to those described for location of veterans in step 5. Of the 830 members of the nonveteran target sample, 667 (80%) were located successfully. (Three were found to be deceased.) 10. The survey questionnaire was mailed to all located potential respondents (599 Vietnam veterans, 566 non-Vietnam veterans, and 667 nonveterans). The original mailing was followed by three other mailings, each 3 weeks apart, and by telephone calls to target respondents that urged them to participate. We offered to interview by telephone those target respondents who did not want to take the time to fill out a pencil-and-paper questionnaire; 15 respondents took us up on our offer. 11. In all, 481 Vietnam veterans, 502 non-Vietnam veterans, and 487 nonveterans returned a completed questionnaire, 10 too late for inclusion in the data analysis
. These numbers were 80%, 89%, and 73% of the respective located target groups to whom a questionnaire had been mailed. We are not sure why the questionnaire return rate for the nonveteran group was lower than that for the two veteran groups. Three explanations appear possible: (a) Our criterion for location was lower for the nonveteran group than for the veteran groups. For the latter groups (because we needed to know whether each man had served in Vietnam), we did not consider a man located until we had spoken in person to him or to a member of his household and obtained information on where he had served. For the nonveteran group, we considered a man located when an outside source--for example, a Department of Motor Vehicles representative or a telephone operator--could give us an address considered current. It is possible that some of these addresses were not really current and that some of our nonveterans never received the survey questionnaire, (b) The fact that we had personal contact with every veteran, or at least with a member of his household, prior to the mailing of the survey questionnaire undoubtedly increased veterans' inclination to take the time tofillout the questionnaire, (c) Because the investigation was called a study of the personal consequences of military service in the Vietnam era, veterans had a greater personal stake in contributing to its success. Several nonveterans could not understand why a study like ours should include them since they never had served. Their puzzlement persisted even after we made numerous attempts to explain the importance of a control or comparison group. At any rate, the response rates obtained for each of the groups were acceptably, even impressively, high. Development of Case Weights Two sets of weights were developed for each case in the final study sample. The first set made Vietnam veterans, non-Vietnam veterans, and nonveterans in the study sample representative, respectively, of all Vietnam veterans, non-Vietnam veterans, and
Epidemiology of PTSD
nonveterans in the ninth-grade class of I960. These weights were used for analyses to answer questions (outside the scope of the present paper) on the antecedents of Vietnamera service: Who served? Who fought? The second set of weights, used for analyses reported in this paper and other analyses on the consequences of Vietnam-era service, made the non-Vietnam veterans and nonveterans similar to the Vietnam veteran group in terms of 51 preservice characteristics. These weights, which produced the matched samples, were used for answering questions that dealt with the consequences of Vietnamera military service and combat. For the Vietnam veteran sample, weights 1 and 2 were identical. To assess the adequacy of the first (population) set of weights, we applied the set to our data and then compared our study sample's mean and standard deviation on two key variables--socioeconomic status and general academic aptitude--with corresponding weighted statistics obtained from the entire sample of 50,000 Project TALENT ninthgrade males. For our study sample of 1,470 men, the weighted mean and standard deviation on the socioeconomic status score were 96.37 and 10.70, respectively; for general academic aptitude, they were 434.31 and 115.5. TALENT population figures were 96.45, 10.49, 430.2, and 119.5. These figures show that our first set of case weights was successful in capturing average populationfiguresto within .03 SD, an impressive accuracy level. To assess the adequacy of the second (matching) set of weights, we applied the set to our data and then compared the three study samples in terms of their mean scores on 51 preservice characteristics measured in the ninth grade. There were no statistically significant
differences on any of these variables. We conclude that our second set of case weights was successful in producing samples of Vietnam veterans, non-Vietnam veterans, and nonveterans well matched on early, preservice characteristics. Measurement of PTSD The American Psychiatric Association's Diagnostic and Statistical Manual (American Psychiatric Association, 1980) lists four indicators of PTSD: 1. "Exposure to recognizable stressor or trauma." 2. "Reexperiencing of trauma through flashbacks, nightmares, or intrusive memories." 3. "Emotional numbing to or withdrawal from external environment." 4. "The experience of at least two symptoms from a list including hyperalertness, sleep disturbance, survival guilt, memory impairment, and avoidance of situations that may elicit traumatic recollections." The 1981 questionnaire was designed to measure general well-being, stress, and functioning. It was not designed specifically to measure PTSD. Therefore, only a rough approximation of the DSM III definition was possible. We searched our questionnaire for indicators of each of the four defining characteristics of the disorder. While we could not come up with a completely satisfactory operational measure of the disorder, we obtained sufficient information in our 1981 questionnaire to reach a crude diagnosis. First, exposure to recognizable stressor or trauma: Proportionately more Vietnam veterans than non-Vietnam veterans or nonveterans were exposed to a recognizable stressor or trauma: Live combat. In addition, it is at least theoretically possible that overseas military service, even outside a war zone, was traumatic for some non-Vietnam veterans. McDermotl (1981) proposes that overseas service
could be associated with a variety of stressors, such as a new and deficient diet, housing, and medical care; extreme weather conditions; loneliness; culture shock; and so forth. If the McDermott hypothesis is correct, then the incidence and seventy of PTSD in the non-Vietnam veteran group should fall somewhere between the Vietnam veteran and non veteran groups. We
Journal of Clinical Psychology, January 1987, Vol. 43, No. I
used membership in the three Study Group
s as a rough indicator of degree of exposure to a recognizable stressor or trauma. Second, reexperiencing of the trauma through flashbacks, nightmares, or intrusive memories: The survey questionnaire contained a direct question on the frequency of nightmares in the past year. There were no direct questions about problems related to flashbacks or intrusive memories; however, there were three items related to what might be called loss of control, a concept that has some overlap with flashbacks and intrusive memories. The items referred to the amount of distress in the past year caused by wanting to break or destroy something, by feeling fearful or apprehensive for no apparent reason, and by being unable to control one's temper. Answers to these four questions were treated as indicators of reexperiencing of the trauma. To be scored positive for PTSD, the respondent had to report at least two of the four symptoms in this group. Third, emotional numbing to or withdrawal from the external environment: There were two indicators of emotional numbing and two indicators of withdrawal from the environment available from the questionnaire--the amount of distress in the past year caused by being unable to get excited about things, the extent to which having someone to be emotionally close to was a problem in the past year, the amount of distress in the past year caused by the feeling that life was not worth living, and a self-rating of one's interest in being with people, relative to other American men of one's age. Answers to these four items were treated as indicators of emotional numbing and withdrawal. To be scored positive for PTSD, the respondent had to report at least two of the four symptoms in this group. Fourth, PTSD-related symptoms, including hyperalertness, sleep disturbance, survival guilt, memory impairment, and avoidance of situations that may elicit traumatic recollections: No questionnaire items dealt with survivor guilt, memory impairment, or avoidance of situations that elicited traumatic recollections. Five items did concern hyperalertness or excessive jumpiness, and three dealt with sleep disturbance. These items were, respectively, the amount of distress in the past year caused by feeling restless or jittery, by letting little things make one angry, by feeling nervous, by feeling tense, and being easily startled by random noises; and the frequency during the past year of episodes in which one had difficulty sleeping, in which one's sleep was disturbed during the night by troubling thoughts, and in which one had serious difficulty getting up in the morning. Answers to these eight items were treated as indicators of the fourth set of PTSDrelated symptoms. To be scored positive for PTSD, the respondent had to report at least two of the eight symptoms in this group. Two PTSD scores were developed for each respondent: First, a crude PTSDdiagnosis score, in which 0 indicated the absence of the disorder and 1 the presence of the disorder (at least two reported symptoms from each of the symptom groups described above); second, a continuous PTSD scale score that ranged from 16 to 80 and indicated the severity of the disorder in each respondent. The dichotomous PTSD score should be treated with caution. We use it as a rough indicator of the presence of PTSD. However, our diagnosis was based only on the reported presence of an imperfect list of symptoms in a research setting, as opposed to the established presence of a more precise set of symptoms in a diagnostic setting. The continuous PTSD scale score had a coefficient alpha reliability of .84. Thus, the scale was internally consistent or homogeneous and probably was measuring a single disorder.
RESULTS Group Differences in the Incidence and Severity of PTSD Nineteen percent of Vietnam veterans had post-traumatic stress disorder at age 36 according to our dichotomous index. (See Table 1.) The corresponding figure for the
Epidemiology of PTSD
other two groups (matched with the Vietnam veterans in terms of 51 characteristics measured in high school) was one-third less (12%). These group differences were statistically significant. The average PTSD scale scores for the three matched groups of men were 29.9, 27.1, and 27.3, respectively. These differences were also statistically significant. Vietnam veterans had the highest PTSD score. Contrary to the McDermott hypothesis, there was no difference in the average PTSD score of the matched nonVietnam veteran and nonveteran groups.
Table 1 Group Differences in PTSD, 1981
Index of PTSD Average score, PTSD scale" Classified as positive for PTSD (at least two symptoms positive on subscales I and 2 and 3) At least one symptom positive on PTSD subscale I (nightmares, panic attacks) At least one symptom positive on PTSD subscale 2 (emotional numbing, withdrawal from external environmeni) At least one symptom positive on PTSD subscale 3 (hyperalertness, excessive jumpiness, disturbed sleep) "Range, 16 to 80; SD = 9.67. *p < .05. **p < .01. ···p < .001.
veterans veterans Nonveterans significant?
The principal PTSD indicators that separated the Vietnam veteran group from the other two matched groups were those that dealt with reexperiencing the trauma (nightmares, loss of control) and those that dealt with emotional numbing and withdrawal from the external environment. Differences in the proportion within each group reporting problems in these areas were quite significant. In contrast, group differences in the other PTSD symptoms (hyperalertness, excessive jumpiness, disturbed sleep) was not significant. Additional statistical tests confirmed that the Vietnam veteran group was indeed significantly different from each of the other two matched groups in terms of PTSD severity, while the other two groups did not differ from each other on this variable. We infer from the pattern behind these results that, in keeping with the residual stress model of Figley (1978) and contrary to the stress evaporation model of Worthington (1977, 1978), PTSD problems that stem from the Vietnam conflict have persisted more than a decade after the traumatic experience. Furthermore, increases in the baseline incidence and severity of PTSD are associated with the Vietnam experience, but not with military service in general. Because of the latter conclusion, all further analyses of the antecedents and correlates of the disorder were confined to the sample of Vietnam veterans. Antecedents of PTSD Among Vietnam Veterans There has been some disagreement about the relative contribution of background characteristics, general military adjustment, and the combat experience to the onset and persistence of PTSD. Proponents of the background or stress predisposition model claim
Journal of Clinical Psychology, January 1987, Vol. 43, No. I
hat those who succumbed to the disorder had the predisposition to be broken by stress :ven before they entered the military. To a large extent, proponents of this background nodel also support the stress evaporation model, which maintains that most of the stress issociated with military service evaporates with the passage of time (Borus, 1974; Carr, 1975; Enzie, Sawyer, & Montgomery, 1973; Strange, 1974; Worthington, 1977, 1978). Their major evidential support for this position is that they have failed to find signifi:ant differences in postservice difficulties between soldiers who were assigned to Vietlam and their contemporaries who were not. In addition, within the group of Vietnam /eterans, they have failed to find differences in postservice difficulties between veterans .vho saw heavy combat and veterans who did not. Finally, they have found that when lelp-seeking veterans are studied, their difficulties are at least as related to preservice difficulties as to the service or combat experiences. In general, this first group of studies has been hampered by a focus on very small, often self-selected, help-seeking samples, generally with a poorly chosen (and occasionally no) control group. On the other side of the argument are those who claim that it is primarily the combat experience itself that triggers the onset of PTSD and that the disorder is long lasting: Residual stress flows from combat many years after the combat experience has passed (Barret-Ruger & Lammers, 1981; Figley, 1978; Figley & Eisenhart, 1975; Kadushin, Boulanger, & Martin, 1981; Laufer, Yager, Frey-Wouters, Donnellan, Gallops, & Stenbeck, 1981; McDermott, 1981; Penk, Robinowitz, Roberts, Patterson, Dolan, & Atkins, 1981; Strayer & Ellenhorn, 1975). In support of their claims, these investigators cite research data that show that Vietnam veterans are more stressed than non-Vietnam veterans (a finding corroborated by our data in Table 1) and that, within the group of Vietnam veterans, those who saw heavy combat are more stressed than those who saw light combat or no combat (a finding corroborated by our data to be presented below). In general, these investigators have failed to find a relationship between preservice or service adjustment and postservice difficulties, although the Kadushin and Laufer investigations found that certain elements of an individual's background (such as being white and coming from a stable family) serve to attenuate the negative effect of combat. Studies in this second group generally were conducted at a later time period than studies in the first group; they also have utilized more sophisticated research design
s and sampling methods. Our data support the second group's point of view. Table 2 presents the relationship between the presence of PTSD at age 36 among Vietnam veterans in our sample and (1) a set of 14 demographic, ability, and personalitybackground characteristics; (2) a set of 5 indices of military adjustment; (3) a set of 10 other (noncombat) variables related to service; and (4) a set of combat-related experiences. Why do some Vietnam veterans develop PTSD, while others do not? The data in Table 2 indicate that the intensity of the combat experience is a strong contributory factor to which veterans will come down with the disorder after the war is over and how severe the disorder will be. Of the 14 background characteristics, 5 indices of military adjustment, and 6 non-combat service behaviors listed in Table 2, only 2--low selfconfidence at age 15 and heavy liquor consumption during the period of military service--were associated significantly with PTSD among Vietnam veterans at age 36. It is impossible to determine from our data whether heavy liquor consumption was a precursor of PTSD (a behavior that indicates susceptibility to breaking down under stress) or a consequence of ongoing combat-related stress problems. In sharp contrast to the lack of association obtained with the background and military-adjustment variables, significant associations were found between the combatexperience variables and PTSD. Ten combat experiences were studied for their possible association with subsequent PTSD. (See the last 10 lines of Table 2.) Each, without exception, was related to the disorder; the more severe the soldier's exposure to combat and injury during the Vietnam war, the greater the number of PTSD-related problems
Epidemiology of PTSD
Table 2 Antecedents of PTSD Among Vietnam Veterans
Hypothesized antecedents of PTSD
Was variable significantly related to PTSD at age 36?
Direction associated with high PTSD
Individual background characteristics, age IS Sociodemographic characteristics Race Socioeconomic status of parents Size of community of origin Academic ability Personality traits Interest in being with people Sensitivity to other people's needs Impulsiveness Energy, vigor Calmness Tidiness, neatness Interest in cultural activities
Leadership capacity Self-confidence Mature personality Indices of military adjustment Amount of liquor consumption during service Amount of drug use during service disciplinary action
during service Military awards Type of military discharge Other (noncombat) variables related to service Military rank Time spent in Vietnam Perceived cohesiveness of military' unit Amount of trust in commissioned officers Amount of trust in noncommissioned officer
s Amount of trust in enlisted men Vietnam combat experience Receive fire from enemy Fire own weapon at enemy Kill enemySee someone get killed See enemy wounded See American wounded See enemy dead See American dead Find self in combat situation where survival was in jeopardy Receive injury
No No No No No No No No No No No No Yes* No Yes* No No No No No No No No No No Yes" a Yes* Yes* Yes* Yes* Yes** a Yes** Yes*
Low self-confidence High liquor consumption Frequent exposure Frequent exposure Frequent exposure Frequent exposure Frequent exposure Frequent exposure Frequent exposure Frequent exposure Frequent exposure Injury requiring hospitalization
'This relationship with PTSD barely missed significance at the .05 level. It was significant at the . 10 level *p < .05. **p < .01. **«p < .001.
as a civilian citizen more than a decade later. We divided the Vietnam-veteran sample into those who had seen relatively heavy combat (those who scored in the top third of a combat scale that consisted of the first nine combat experiences) and those who had not. Incidence of PTSD was 27% in the former group, afiguresignificantly higher than either the 19%figurein the Vietnam-veteran group taken as a whole or the 12% baseline
Journal of Clinical Psychology, January 1987, Vol. 43, No. I
figure in the groups of non-Vietnam veterans and nonveterans matched with the Vietnam group on 51 high school characteristics. The data, while correlational, are consistent with the conclusion that heavy combat has long-lasting effects on psychological health and imposes lingering stress on many soldiers' lives long after the combat experience. Sociodemographic, Health, and Social Correlates of PTSD Among Vietnam Veterans How does PTSD affect other spheres of human functioning? We looked at the degree of association between PTSD and a gamut of indicators of performance in other dimensions of living, which included marital status, educational achievement, occupational achievement, absence from work, substance use and abuse, brushes with the law, contact with a counselor for help with problems, hospitalization for other-than-physical problems, and the Vietnam veteran's evaluation of the overall effects of military service on his subsequent life. Results are given in Table 3. The presence of a significant association in the table does not imply that PTSD caused the problem because correlation does not imply causality. The relation between PTSD and its correlates is likely to be one of reciprocal causality. For example, problems with one's spouse or work caused by PTSD exacerbate PTSD, which, in turn, leads to further problems with the family and, thus, contributes to the persistence of the man's problems through time. Table 3 shows that a high PTSD score is associated significantly with living alone and with divorced, separated, and single marital status. As previously stated, the direction of causality is not clear from our data. Does PTSD lead to marital break-up, or does living alone heighten PTSD-related problems? In all likelihood, both forces arc at work. PTSD was related very slightly to low educational attainment and was unrelated to occupational attainment. There was a clear and strong association between the presence of PTSD and the presence of other health and social problems. A high PTSD score was associated with missed days at work, relatively heavy drug consumption, relatively more frequent arrests, with seeking professional help for problems, and with hospitalization for other-than-physical problems. PTSD also was associated strongly with the perception that military service had negative instead of positive effects on one's future life. We conclude that the effects of PTSD spill over into other areas of human functioning in a manner that is inevitably unfavorable for many Vietnam veterans' quality of life.
Variables That Moderate the Incidence or Severity of PTSD among Vietnam Veterans Our national study investigated whether certain elements of the Vietnam veterans' environment were associated with reduced levels of PTSD. Another large national study of Vietnam veterans (Laufer et al., 1981) found that the association between combat and postservice stress was "confined mainly to veterans who served between 1968 and 1974" (p. 315), the latter years of the war. Separate analyses of the Laufer et al. data (Kadushin et al., 1981) also found that the presence of friends who are also Vietnam veterans "helps to reduce current levels of stress" (p. 477). Our 1981 questionnaire had information on when each Vietnam veteran served in Vietnam. Although we did not have information on the veterans' present friendship networks, we did have information on some potentially related aspects of social support
: The size of the community in which the veteran currently was residing and the degree to which the veteran actively participated in five types of organizations: Church, social, civic, professional, and charitable. Wc performed a series of analyses of variance with these timing and environmental variables as independent variables
, PTSD in its dichotomous and continuous form as
Epidemiology of PTSD
Table 3 Sociodemographic, Health, and Social Correlates of PTSD Among Vietnam Veterans
Hypothesized correlate of PTSD (measured concurrently at age 36) Sociodemographic status Demographic characteristics Marital status Living with wife or girlfriend Educational achievement High-school degree College degree Number of years of schooling completed Occupational achievement Job prestige Hourly pay Yearly pay Health and social problems Absence from work Days ill Substance use and abuse Cigarette consumption Liquor consumption Drug consumption Pain medication Tranquilizers Sedatives, sleeping pills Prescription stimulants Nonprescription stimulants Narcotics Marijuana and street drugs Brushes with the law Traffic tickets Arrests Felonies Contact with counselor for problems Marital or job adjustment Alcohol or drug use Nervous condition Other mental health problem
Hospitalization for problems Marital or job adjustment Alcohol or drug use Nervous condition Other mental health problem Evaluation of overall (positive or negative) effects of service
Was variable significantly related to Direction associated with PTSD at 36? high PTSD
Yes** Yes" No No * No No No
Divorced, separated, single Not living with wife or girlfriend Low educational attainment
Yes** No No No " a Yes*" Yes** Yes* Yes* No Yes* ' No Yes* Yes*** Yes* Yes* Yes* No No Yes**
Relatively high absenteeism Relatively heavy consumption Relatively heavy consumption Relatively heavy consumption Relatively heavy consumption Relatively heavy consumption Relatively heavy consumption Having been arrested Having been convicted Having sought help for this Having sought help for this Having sought help for this Having been hospitalized for this Having been hospitalized for this Perception that service had negative effects for one's future
aThis result barely missed significance at the .05 level. It was significant at the .10 level. *p < .05. **p < .01. * " p < .001.
Journal of Clinical Psychology, January 1987, Vol. 43, No. I
the dependent variable, and amount of combat experienced in Vietnam as a covariate. We found only participation in church activities to be associated with reduced incidence and severity of PTSD.
SUMMARY AND CONCLUSION At age 36, Vietnam veterans in the high school class of 1963 reported significantly more problems related to nightmares, loss of control over behavior, emotional numbing, withdrawal from the external environment, hyperalertness, anxiety, and depression than their classmates matched with them on 51 high school characteristics. These problems correspond closely to the disorder labeled post-traumatic stress disorder (PTSD) by the American Psychiatric Association (1980). Our index of the incidence of PTSD showed 19% of Vietnam veterans to be suffering from the disorder at age 36. The corresponding figure for matched groups of non-Vietnam veterans and nonveterans was significantly different: 12%. Among Vietnam veterans, the primary antecedent of PTSD was the severity of the combat experience, not the background characteristics of the individual soldier or his general military behavior and adjustment. Twenty-seven percent of Vietnam veterans who experienced heavy combat manifested PTSD-related symptoms at age 36, a figure significantly higher than the 19% baseline figure for Vietnam veterans taken as a whole or the 12% figure for the Vietnam veteran and nonveteran groups. PTSD was associated with other family, mental health, and social interaction problems. Some environmental variables--for example, the presence of a spouse, or being a churchgoer--were associated with reduced levels of PTSD or with reductions in the degree of association between combat and PTSD. The direction of cause and effect in these associations cannot be ascertained from our data, but it seems plausible to postulate that support factors can and do help some Vietnam veterans with PTSD. For the most part, these findings are consistent with those of other large-scale studies of nationally representative groups of Vietnam veterans, non-Vietnam veterans, and nonveterans. Our findings are particularly illuminating because of our ability to match our three samples on pre-military characteristics. We believe that the evidence is overwhelming that heavy combat can and does lead to a syndrome of symptoms indicative of emotional distress and that the combat-induced distress often lasts a long time, certainly more than a decade in many cases. Our research did not focus on post-combatexposure events, behaviors, and treatments that precipitated the disorder or ameliorated it. Future work will investigate these issues.
REFERENCES AMERICAN PSYCHIATRIC ASSOCIATION TASK FORCE ON NOMENCLATURE AND STATISTICS. (1980). Diagnostic and Statistical Manual of Mental Disorders (3rd ed.). Washington: American Psychiatric Press. BARRET-RUGER, D., & LAMMERS, C. A. (1981). Vietnam combat veterans: Assessment of pre-military and military influences on post-military adjustment. Paper presented at the annual meeting of the American Psychological Association
, Los Angeles
. BORUS, J. F. (1974). Incidence of maladjustment in Vietnam returnees. Archives of General Psychiatry. 30, 554-557. CARD, J. J. (1983). Lives after Vietnam: The personal impact of military service. Lexington, MA: Heath. CARR, R. A. (1973). A comparison of self concept and expectations concerning control between Vietnam- era veterans and non-veterans. Ph.D. dissertation, St. Louis University. ENZIE, R. R., SAWYER, R. N., & MONTGOMERY, F. A. (1973). Manifest anxiety o f Vietnam returnees and undergraduates. Psychological Reports, 33, 446. FIGLEY, C. R. ( E D . ) . (1978). Stress disorders among Vietnam veterans. New York: Brunner/Mazel. FIGLEY, C. R,, & EISENHART, W. (1975, August). Contrasts between combat and non-combat Vietnam veterans regarding selected indices of interpersonal adjustment. Paper presented at the annual meeting of the American Sociological Association, San Francisco
Epidemiology of PTSD
KADUSION, C,, BOULANGER, G., & Martin, J.
(1981). Long-term stress reactions: Some causes, consequences, and naturally occurring support systems. Legacies of Vietnam: Comparative adjustment of veterans and their peers (Vol. 4). Washington: Government Printing Office. LAUFER, R. S., YAGER, T., FREY-WOUTERS, E., DONNEUAN, J,, GALLOPS, M., & STENBECK, K. (1981). Post-war trauma: Social and psychological problems of Vietnam veterans in the aftermath of the Vietnam War. Legacies of Vietnam: Comparative adjustment of veterans and theirpeers (Vol. 3). Washington: Government Printing Office. MCDERMOTT, W. F. (1981). The influence of Vietnam combat on subsequent psychopalhology. Paper presented at the annual meeting of the American Psychological Association, Los Angeles. PENK, W. E., ROBINOWITZ, R., ROBERTS, R., PATTERSON, E. T., DOLAN, M. P., & ATKINS, H. G. (1981). Adjustment differences among male substance abusers varying in degree of combat experience in Vietnam. Journal of Consulting and Clinical Psychology, 49, 426-436. STAMPLER, F. M., & SIFPRELLE, R. C. (1981). The psychological adjustment of Vietnam era veterans: The next decade. Paper presented at the annual meeting of the American Psychological Association, Los Angeles. STRANGE, R. E. (1974). Psychiatric perspectives of the Vietnam veteran. Military Medicine, 139, 96-98. STRAYER, R., & ELLENHORN, L. (197S). Vietnam veterans: A study exploring adjustment patterns and attitudes. Journal of Social Issues, 31, 81-91. WORTHINGTON, E. R. (1977). Post-service adjustment and Vietnam era veterans. Military Medicine. 142, 865-866. WORTHINGTON, E. R. (1978). Demographic and pre-service variables as predictors of post-military service adjustment. In C. R. Figley (Ed.), Stress disorders among Vietnam veterans. New York: Brunner/Mazel.