| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLE |
From the National Center for Posttraumatic Stress Disorder, Clinical Neurosciences Division, Veterans Administration Connecticut, and Department of Psychiatry, Yale University School of Medicine, New Haven, Connecticut.
Address reprint requests to: Sheila Wang, PhD, Department of Psychiatry, VA Connecticut/116A, 950 Campbell Ave, West Haven, CT 06516. E-mail: wang.sheila{at}west-haven.va.gov
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHOD: Clinical symptoms were assessed in and serum thyroid measures were obtained from 12 WWII veterans with PTSD and 18 WWII veterans without PTSD.
RESULTS: WWII veterans with combat-related PTSD showed elevations of serum total and free T3 with no elevations of free T4 and TSH compared to control subjects, replicating the results of our previous studies. A significant positive relationship between total and free T3 and PTSD symptoms, specifically hyperarousal symptoms, was also replicated in the total WWII group. Elevations of total T4 and TBG were not replicated in the WWII group with PTSD, which may indicate a shift with age in the free/bound dynamics of the thyroid alterations observed.
CONCLUSIONS: This study supports the observation that the thyroid system is altered in chronic combat-related PTSD. The observed alterations of thyroid function along with PTSD symptoms appear to be chronic, detectable 50 years after the war.
Key Words: posttraumatic stress disorder, • thyroid, • triiodothyronine, • combat, • World War II veterans, • psychiatric symptoms.
Abbreviations: PTSD = posttraumatic stress disorder;; T3 = triiodothyronine;; T4 = thyroxine;; TSH = thyrotropin;; WWII = World War II;; TBG = thyroxine-binding globulin;; PSS = PTSD Symptom Scale;; POW = prisoner of war.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
With this rationale in mind, we added a complete thyroid assessment to our profile of stress-responsive hormonal measures in studying the psychoendocrinology of PTSD in combat veterans. In previous studies, we observed an unusual thyroid profile in these veterans, including elevated total T3, free T3, and total T4, but no elevation of free T4 or TSH compared to control subjects. We have replicated these findings in four groups (N = 96) of Vietnam combat veterans (7) and a group (N = 11) of Israeli combat veterans (8) with PTSD. The thyroid elevations did not typically exceed the normal range, as specified for the diagnosis of glandular disease in the field of clinical endocrinology, but there is evidence that relatively modest changes in thyroid hormone levels may have important clinical significance in relation to psychiatric disorders (9).
In exploring the clinical significance of the T3 elevations in combat-related PTSD, we found significant positive correlations between total T3, free T3, and PTSD symptoms, specifically frequency of hyperarousal symptoms measured by the Clinician-Administered PTSD Scale in a sample (N = 65) of Vietnam veterans (10) and novelty seeking subscale scores on the Cloninger Tridimensional Personality Questionnaire in another sample (N = 27) of Vietnam veterans (11). Many of the symptoms of hyperthyroidism are similar to the hyperarousal symptoms observed in PTSD, for example, irritability, difficulty sleeping, difficulty concentrating, anger outbursts, and exaggerated startle. Because T3 is two to four times more biologically active than T4, the significant positive correlation between T3 and hyperarousal seemed to provide evidence of a potentially important hormone-symptom relationship in this disorder. To determine whether these findings could be replicated and generalized more broadly in combat-related PTSD, we studied WWII veterans to investigate whether their biological characteristics, 50 years after the war, reflected the thyroid alterations and the hormone-symptom relationships we observed in younger Vietnam veterans.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
Hormonal Samples
Blood samples (10 ml) in red-topped (untreated) vacuum tubes for thyroid hormone assays were collected between 8 and 9 AM in 27 of the 30 subjects; three blood samples were collected later in the day. After setting of the clot and centrifugation, the serum was divided into three 1.5-ml aliquots in small glass vials and frozen at -70°C until assayed. Because six different hormonal assays were to be performed on each sample, the three aliquots minimized freezing and rethawing cycles as a potential source of hormonal instability and analytic error, especially since two different hormonal assays were usually done concurrently when each aliquot was thawed.
Serum total T4, free T4, total T3, and TBG concentrations were measured by radioimmunoassay (RIA) procedures using commercially available kits (Incstar Corp., Stillwater, MN). The interassay coefficient of variation in our laboratory was 3.7% for total T4, 4.2% for free T4, 6% for total T3, and 4.0% for TBG. Serum-free T3 concentrations were measured using an RIA kit procedure (Diagnostic Products Corp., Los Angeles, CA). The interassay coefficient of variation in our laboratory was 2.7% for free T3. Serum TSH concentrations were measured by means of a sensitive third-generation immunoradiometric procedure (Incstar Corp.), and the interassay coefficient of variation was 4.0% in our laboratory.
Clinical Measures
The following clinical measures were administered to assess PTSD symptoms, combat exposure, and general psychiatric symptomatology: the Mississippi Scale for Combat-Related PTSD (12); the PTSD Symptom Scale (PSS) (13), which includes symptom cluster subscales; the Combat Exposure Scale (14); and the Brief Symptom Inventory (15). As stated above, the criterion for a diagnosis of PTSD was a Mississippi Scale score of 107 or above. Age, height, weight, years of education, medical problems, medications, and history of or current substance abuse, smoking, and suicidality were obtained during an interview with each subject. When possible, information was confirmed by hospital records.
Data Analysis
All thyroid measures were included in an overall one-factor multivariate analysis of variance to determine whether there were overall mean differences between the two groups when all dependent variables were considered simultaneously. Subsequent univariate t tests were performed on each dependent variable. Pearson product-moment correlations were used for correlational analyses. On the basis of our previous work, we predicted higher thyroid hormone levels in the PTSD group and positive correlations between thyroid measures and clinical measures. Therefore, we used one-tailed probability values for all t tests and individual correlation coefficients. Bonferroni probability values were calculated to correct for the number of correlations in the correlation matrix.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
< .05) for correlations between thyroid measures and clinical measures of PTSD are shown in Table 2. Correlations that are also significant after correction for the number of correlations in the matrix using the Bonferroni procedure are indicated. The strength of the positive correlation between both total and free T3 and hyperarousal symptoms, measured by the PSS, is supported by a finding of significance after correction for multiple correlations.
|
Serum-Free T3
Table 1 shows significantly higher mean levels of serum-free T3 in WWII veterans with PTSD compared to WWII veterans without PTSD (3.45 vs. 3.01 pg/ml, t = 3.38, p < .001). Significant positive individual correlations were found between serum-free T3 and each clinical PTSD measure administered: the Mississippi Scale score, the PSS total score, and each PSS subscale (reexperiencing, avoidance, and hyperarousal). After Bonferroni corrections were made, only the reexperiencing subscale of the PSS failed to reach significance at the .05 level. Highly significant positive correlations were observed between serum-free T3 and both the hyperarousal subscale of the PSS (Figure 1; r = .60, p < .0003) and total Mississippi Scale scores (r = .61, p < .0002). Both correlations were significant at the .01 level after Bonferroni correction for multiplicity. The individual correlation between serum-freeT3 and the Combat Exposure Scale score was comparatively weak and of borderline positive significance (r =.31, p < .06).
|
Total T3/Free T4 Ratio
Previously, we hypothesized (6) that because about 80% of the body’s supply of T3 is produced by peripheral conversion of free T4 to T3, elevations of free and total T3 found in combat-related PTSD patients might reflect increased peripheral conversion and that a useful indicator of the rate of the conversion process might be the total T3/free T4 ratio. A higher ratio would accordingly represent increased conversion of free T4 to T3. A significantly higher total T3/free T4 ratio (138 vs. 115, t = 2.32, p < .014) was found in WWII veterans with PTSD compared to WWII veterans without PTSD, supporting the hypothesis of increased conversion of free T4 to T3.
Serum TBG
No significant mean difference in serum TBG was found between the PTSD group and the comparison group (31.7 vs. 28.7 µg/ml, t = 1.16, p < .13). Both groups had relatively high levels of TBG (reference range: 12–30 µg/ml). Significant individual correlations between TBG and the Mississippi Scale score and the PSS hyperarousal subscale were found (Table 2; r = .31, p < .05 and r = .35, p < .03, respectively); however, after Bonferroni correction, no significant correlations were observed between TBG and any clinical measures of PTSD.
Serum TSH
No significant mean difference in serum TSH between the two groups was observed, although there was a trend toward higher TSH in the PTSD group (2.21 vs. 1.46 µIU, t = 1.53, p < .07). This could indicate the contribution of increased central drive to the elevations of T3 in addition to the hypothesized augmented peripheral conversion of T4 to T3 in this population. In our previous studies with Vietnam and Israeli veterans, we did not observe a trend toward higher TSH. The individual correlation between the Mississippi Scale score and TSH was significant (r = .32, p < .04); however, it failed to reach significance after Bonferroni correction.
Clinical Measures
WWII veterans with PTSD reported significantly more symptoms on every clinical measure administered and almost twice the amount of combat exposure as WWII veterans without PTSD. The level of combat in the comparison group was light to moderate and in the PTSD group was moderate to heavy. Because the criteria for dividing the two groups was based on a score of 107 or more on the Mississippi Scale, it was expected that the other PTSD measures would also be quite different in the two groups. Scores from the Brief Symptom Inventory, which was designed to reflect general psychological symptom status, were also significantly elevated in the PTSD group (mean General Severity Index, 2.5 vs. 1.0, t = 5.6, p < .001). The subscales of the Brief Symptom Inventory include somatization, obsessive-compulsive behavior, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, psychoticism, and additional items. This finding indicates that chronic PTSD does not simply result in elevated core symptoms associated with PTSD but seems to be related to a broad range of psychological symptoms.
Suicidality
History of suicidal thoughts was reported significantly more frequently in the PTSD group than in the comparison group (67% vs. 22%, 
2 = 5.93, p = .015).
Age, Height, Weight, Years of Education, Medical Problems, Substance Abuse, and Smoking
There were no significant mean differences in age (71.4 vs. 72 years), height (68 vs. 69 inches), weight (187 vs. 191 lb), or years of education (12.2 vs. 12.1 years) between the PTSD and comparison groups. Chi-square analyses of frequency of medical problems based on several categories (asthma, hypertension, diabetes, cardiovascular incident, heart disease, ulcer, cancer, emphysema, gastrointestinal problems, and miscellaneous medical problems) showed no significant differences in frequency of these medical conditions between the two groups. No differences (frequency) in history of or current substance abuse or smoking were found between the PTSD and comparison groups.
Medications
Most of the subjects in both groups were taking various medications. For statistical comparison, medications were grouped into three categories: psychiatric, cardiovascular, and other medications. Chi-square analyses showed no significant differences in frequency of prescription medications between the two groups in any category.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
The normal range for serum total T3 is 70 to 190 ng/dl (17). In our previous thyroid study (7), the control group (mean age = 38 years, N = 24) value for total T3 was 127 ± 24 ng/dl. Considering the observation that T3 tends to decrease with age, the WWII veterans in both the PTSD and comparison groups in the present study had marked elevations (177 and 152 ng/dl, respectively). Given the level of combat and perhaps partial PTSD in our comparison group, an age-matched civilian control group might reveal an even more dramatic elevation of T3 in WWII veterans with PTSD.
Elevations of total T4 and TBG previously observed in younger combat veterans with PTSD (6) were not observed in the older WWII PTSD group. This difference might be explained by the increased production of TBG observed in the elderly (18–19). Because TBG levels were elevated in both elderly groups and because TBG has a higher affinity for T4, it follows that the differences in total T4 between the PTSD and comparison groups would be reduced. To a lesser extent, because TBG has a lower affinity for T3, the same dynamic holds true for total T3. Therefore, the increased TBG production due to aging might have concealed some of the differences in bound thyroid hormones between the two elderly groups. Our findings of more significant elevations of free T3 in older combat veterans with PTSD and more significant elevations of total T3 in younger Vietnam combat veterans with PTSD support this notion. A similar pattern emerges in the T3-symptom correlations, with free T3 being most significantly related to symptoms in the older group and total T3 being most significantly related to symptoms in younger combat veterans.
The total T3/free T4 ratio was significantly elevated in the WWII PTSD group, replicating our earlier finding in Vietnam veterans with PTSD and providing additional support for the hypothesis of increased peripheral conversion of T4 to T3 in combat-related PTSD (7). In contrast to the Vietnam veterans with PTSD, the WWII PTSD group had a nonsignificant trend toward higher TSH compared with control subjects, which may suggest that central nervous system drive, in addition to peripheral conversion, may contribute to the elevated T3 measures observed in the older veterans.
The consistent and robust elevations in T3 we have observed in veterans with combat-related PTSD were probably not detected in routine clinical thyroid function tests because, in general, they do not include direct measures of T3. Although elevations of T3 in this group are largely still within the normal range as defined in the field of clinical endocrinology and do not indicate glandular pathology, the strong positive correlations between T3 and PTSD symptoms, specifically hyperarousal symptoms, seem to point to a potentially clinically significant hormone-symptom relationship. Replication of this relationship in two different groups of combat veterans with PTSD a full generation apart may indicate a need for further study of the clinical importance of T3 levels in this population.
The question of whether symptoms occur in response to higher levels of T3 or whether higher levels of T3 occur in response to increased symptoms indicates a need for placebo-controlled pharmacologic studies whereby T3 is lowered and clinical symptoms are monitored. The few open trials using propranolol, which lowers T3, to treat PTSD symptoms have reported positive results (20–21). Because the complete role of T3 in relation to PTSD symptoms is not yet clear, pilot studies in which there is careful monitoring of clinical responses to lowering T3 should be completed before a large trial is initiated, because although T3 seems to be related to disturbing PTSD symptoms, elevations of T3 could have an adaptive purpose, perhaps modifying other types of symptoms or physiologic processes.
Clinical Measures
The chronicity of PTSD symptoms as well as other psychological symptoms in WWII veterans due to combat stress has been documented extensively in the literature by investigators at many specific time points [eg, 5 (22), 9 (23), 20 (24), 24 (25), and 50 years after the war (26)] and by several investigators (27–29). Our clinical data support the findings of previous studies of the chronicity of PTSD symptoms in WWII veterans and point to a general increase in overall psychological symptoms in this group.
It has been suggested that the neurobiological changes observed in PTSD may have more to do with exposure to traumatic stress than with PTSD. Although our sample size (N = 30) is somewhat small for correlational analyses, our finding that the probability value for the correlation between free T3 and PTSD symptoms (p < .0002) is 200 times more significant than the probability value for the correlation between free T3 and reported combat exposure (p < .06) seems to indicate that the thyroid alterations observed in these veterans may be more specifically related to the disorder of PTSD than to combat exposure alone.
Specificity of Thyroid Findings in PTSD
The consistency and robustness of the T3 elevations observed in veterans with combat-related PTSD in different regional populations, cultures, and age groups strongly suggests that the thyroid system is significantly altered in this population.
We have observed a specific relationship between combat-related PTSD and elevations in serum total and free T3. PTSD as a result of other traumatic experiences may reveal different thyroid profiles. Our preliminary work with POWs with PTSD suggests that these men do not have elevations in total and free T3. A pilot sample population of five WWII POWs and three Korean War POWs had total and free T3 levels significantly below the control group mean. The POWs’ descriptions of their traumatic experiences and their adaptive responses to those experiences makes apparent that they speak in very different terms compared to combat veterans who were not POWs. For example, combat veterans often talk about intensive fighting or fleeing in response to combat. In contrast, POWs report that a fighting or fleeing strategy would likely get them killed. Instead, they describe a withdrawal strategy in terms of "shutting down" or "stonewalling" as an adaptive response to the traumatic stress of long-term captivity. Similarly, decreased levels of thyroid hormones were reported in a recent study of East German refugees suffering from psychiatric disorders, including PTSD, after exposure to prolonged stress (30). These refugees were subjected to unpredictable acts of repression and persecution by the State Security Police, including frequent summonses, interrogations, imprisonment, surveillance at home and work, and other forms of harassment.
Differences in adaptive responses to traumatic stress, partly determined by environmental constraints, may influence whether the thyroid system is activated or suppressed. It is possible that physiologic responses to a chronic life-threatening situation can elevate thyroid hormones, stimulate the sympathetic-adrenal-medullary fight-or-flight system, and result in resetting of the metabolic system toward mobilization and catabolism. However, if the traumatic events occur in an environment in which the fight-or-flight response is not adaptive for survival (eg, in a POW situation, the Holocaust, an oppressive political situation, or some domestic abuse situations), then a life-threatening stressor could result in an adaptation toward conservation/withdrawal and a resetting of the metabolic system toward conservation, anabolism, and decreased thyroid measures. Henry (31) has discussed the contrast in neuroendocrine profiles as the result of active vs. passive coping strategies in response to perceived threat. More specific attention to the adaptive mechanisms used to survive the trauma may be warranted to more fully understand the role of thyroid hormones in PTSD. This point is important to consider because our data do not suggest that all patients meeting criteria for PTSD will have elevated T3 levels. Additional thyroid studies in different PTSD populations with attention given to both short- and long-term adaptive strategies used in response to traumatic experiences will help to clarify the specificity of thyroid alterations in PTSD.
Limitations
The number of subjects (N = 30) in this study was small for correlational analyses. Perhaps a larger sample might reveal significant associations between thyroid hormones and other measures (eg, the Combat Exposure Scale). Correlational analyses do not imply causality. From our study, there is no way to determine whether combat veterans with PTSD had elevations in T3 before they were exposed to combat trauma and were more vulnerable to developing PTSD symptoms or whether T3 elevations occurred after their combat exposure and consequently were associated with symptoms.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
The current replication of a significant positive relationship between T3 and PTSD symptoms, especially hyperarousal symptoms, provides additional evidence of a potentially important T3-hyperarousal symptom association and may offer a rationale for assessment of T3 and pharmacologic intervention to reduce T3 in combat-related PTSD.
Preliminary data indicate that T3 is not elevated and may be decreased in WWII and Korean War POWs, suggesting that elevations of T3 may be specific to combat-related PTSD. Other PTSD populations exhibiting different adaptations to traumatic experiences, perhaps due to environmental constraints, may also show distinct thyroid profiles.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication April 20, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. J. E. Dirkzwager, P. G. van der Velden, L. Grievink, and C. J. Yzermans Disaster-Related Posttraumatic Stress Disorder and Physical Health Psychosom Med, June 1, 2007; 69(5): 435 - 440. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Girdler, K. S. Thompson, K. C. Light, J. Leserman, C. A. Pedersen, and A. J. Prange Jr. Historical Sexual Abuse and Current Thyroid Axis Profiles in Women With Premenstrual Dysphoric Disorder Psychosom Med, May 1, 2004; 66(3): 403 - 410. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Girdler, A. Sherwood, A. L. Hinderliter, J. Leserman, N. L. Costello, P. A. Straneva, C. A. Pedersen, and K. C. Light Biological Correlates of Abuse in Women with Premenstrual Dysphoric Disorder and Healthy Controls Psychosom Med, September 1, 2003; 65(5): 849 - 856. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Lamprecht and M. Sack Posttraumatic Stress Disorder Revisited Psychosom Med, March 1, 2002; 64(2): 222 - 237. [Abstract] [Full Text]
|
|
|
|
|
|
A. J. Prange Jr. Thyroid Axis Sustaining Hypothesis of Posttraumatic Stress Disorder Psychosom Med, March 1, 1999; 61(2): 139 - 140. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
