Mixed-Handedness in Identical Twins Discordant for Combat Exposure in Vietnam: Relationship to Posttraumatic Stress Disorder

There is a growing body of research investigating the relationship between handedness and psychiatric conditions. Studies have shown that non–right-handedness is associated with attention deficit hyperactivity disorder,¹ psychosis,² autism spectrum disorder,³ and intellectual disability,⁴ among other conditions. Mixed-handedness has been associated with greater peritraumatic emotional distress during exposure to traumatic events,⁵ greater prevalence of posttraumatic stress disorder (PTSD) in adults⁶ as well as in children,⁷ and poorer response of PTSD to psychosocial treatments.⁸ Studies have found an increase in mixed-handedness in mainly right-handed veterans with PTSD (i.e., PTSD veterans were not as strongly right-handed as controls).^6,9 It has also been reported that individuals in the general population who met PTSD criteria were more likely to be left-handed.¹⁰

It may be that lesser right-handedness is a pretrauma vulnerability factor for PTSD. Alternately, neurological changes related to the traumatic event or subsequent PTSD might affect handedness. An acquired subtle weakening of preference for using the right hand may be the source of some of the differences seen in previous research, rather than hand preference acting as a risk factor. Previous research addressing whether an identified characteristic of individuals with PTSD is attributable to a predisposition, or instead represents an acquired trait resulting from the traumatic event or the subsequent PTSD, employed monozygotic twin pairs discordant for trauma (i.e., combat, exposure).¹¹ The combat-unexposed twins of combat veterans with PTSD provide a control for genetic and shared early environmental and developmental influences. In this study, self-reported handedness data in combat-exposed veterans with PTSD and their (high-risk) unexposed, identical cotwins, as well as combat-exposed veterans without PTSD and their (low-risk) unexposed, identical cotwins, were analyzed in an attempt to clarify whether mixed-handedness is a pretrauma risk factor for PTSD (in which case it might well also be present in the high-risk cotwins) or is a trait acquired as a result of the combat exposure or subsequent PTSD (in which case there would be no reason to expect it to be present in the high-risk cotwins, whose handedness should be similar to the low-risk cotwins).

Methods

Subjects

Subjects were drawn from a previously described original pool of 103 male identical twin pairs discordant for combat exposure in Vietnam.¹² A total of 62 pairs returned for participation and were asked to complete the handedness questionnaire. In two of these pairs, questionnaires were missing from both members, and the pair was excluded. In another two of these pairs, questionnaires were missing from one member. In one pair, both members were predominantly left-handed, and the pair was excluded from the analysis. In seven pairs, one member was predominantly left-handed; the data from those subjects were also excluded. Because the statistical program implemented was able to handle missing data, data from the twin of an excluded subject were retained. The final number of subjects was 109. No subject had a known history of traumatic brain injury or other neurological condition.

Psychodiagnostics and Psychometrics

The categorical presence or absence of combat-related PTSD was determined by an experienced, doctoral-level clinician using the Clinician-Administered PTSD Scale (CAPS).^13,14 This instrument also provided a continuous measure of lifetime PTSD symptoms, as shown in Table 1. Severity of combat exposure was quantified by means of a retrospective, self-report scale.¹⁵

Handedness Measure

Each subject completed the self-report Edinburgh Handedness Inventory¹⁶ with regard to their lateralization of hand preference. This instrument contains 10 items describing an action. Subjects rate their preference for using one or the other hand for each item. Subjects can respond no preference, left or right preference, or strong left or strong right preference if they prefer one hand exclusively for the task. Typically, the score is calculated by tallying the number of responses for right and for left. Each strong response is given 2 points. For example, if an individual chose right for eight tasks (8 points), left for one task (1 point), and strong right for one task (2 points), he or she would have 10 points right and 1 point left. To calculate the subjects’ handedness, the left points are subtracted from the right points (in this example, 10–1=9), and the resulting score is then divided by the total amount of points (in this example, 11). The resulting handedness score is on a scale ranging from −100, which indicates a preference of the left hand over the right across all tasks, to 100, which indicates a preference of the right hand over the left hand across all tasks. In the previous example, the individual would have received a score of 81.8.

However, the above scoring method discounts how strongly the individual completing the inventory prefers to use a particular hand. If someone prefers his or her right hand for every task but will occasionally use the left hand on some tasks, the individual will still score 100, which is the same score another person would get if he or she preferred the right hand for every task and never used the left hand for any task (“pure” right-handed.) However, these two hypothetical individuals do not, in actuality, have the same degree of hand preference or lateralization. Accordingly, instead of using this standard method for scoring the Edinburgh Handedness Inventory, we implemented a modified scoring system of our own devising. Handedness was scored in relationship to the total possible right-handed choices (all 10 strong right responses). On this modified scale, a score of 100 would indicate an exclusive preference for using the right hand (10 strong right responses), as on the standard scoring system. However, an individual who had a constant but nonexclusive preference (10 right responses) for using the right hand on each task would only score a 50 on this scale instead of the 100 he or she would score using the standard system. A score of 0 would still indicate no preference (i.e., “pure” mixed-handedness). Scores under 0 would indicate greater left- than right-handedness. Thus, this modified scoring system enhances discrimination among individuals with subtler differences in hand preference.

We focused on the degree of lateral preference among the right-handed subjects only. Data from predominantly left-handed subjects were excluded due to their small number and the consideration that left-handedness may be qualitatively different from mixed-handedness.

Data Analysis

The data were analyzed using a mixed model for repeated measures that treated PTSD diagnosis in the combat-exposed twin as a between-pairs fixed effect, combat exposure as a within-pairs fixed effect, and twin pairs as a random effect.¹⁷ If a dependent variable represents a PTSD vulnerability factor, the model predicts a significant diagnosis main effect in the absence of a diagnosis × exposure interaction. If a dependent variable represents a sign acquired solely as a result of trauma exposure, the model predicts an exposure main effect. If a dependent variable represents an acquired PTSD sign, the model predicts a significant interaction. The dependent measure of greatest interest was the modified handedness variable previously described, on which a higher score indicates more right-handedness and a lower score indicates more mixed-handedness. Because of the clear directionality of the hypothesis (i.e., PTSD < non-PTSD and combat exposed < combat unexposed), results in the opposite direction were considered nonsignificant, and p values for results in the predicted direction were halved (i.e., one-sided tests were used). For comparison of handedness according to PTSD diagnostic status, independent t tests were performed within combat-exposed and combat-unexposed subjects separately, again with the unidirectional prediction that PTSD < non-PTSD. Independent t tests were also applied to combat severity and lifetime CAPS score in combat-exposed subjects only, in which the directional predictions were PTSD > non-PTSD. Selected analyses were also performed with combat severity included as a covariate. In addition, for reasons that will become apparent below, a bootstrap version of the Sobel test that implements a SAS (Cary, NC) macro¹⁸ (http://www.afhayes.com) was used to test the possibility that combat severity in combat-exposed twins mediated the association between handedness in combat-unexposed cotwins and PTSD in combat-exposed twins.

Results

Table 1 presents the group means, standard deviations, and results of the unadjusted mixed model analyses and t tests. For the handedness variable, there was a significant diagnosis main effect in the absence of a significant exposure main effect or diagnosis × exposure interaction. Right-handedness was lower (i.e., mixed-handedness was greater) in PTSD than in non-PTSD pairs. An independent t test revealed a significant difference between high-risk and low-risk, combat-unexposed cotwins, with high-risk cotwins showing less right-handedness and a group difference of 9.7 points. An independent t test between PTSD and non-PTSD combat-exposed twins did not yield significant results, but the 95% confidence limit for the group difference was 10.2 points in the predicted direction.

The significant difference between the high-risk and low-risk cotwins was confounded by a significant correlation between the unexposed cotwins’ handedness scores and their combat-exposed brothers’ combat severity scores (r=−0.29, N=57, p=0.03). Analysis of covariance adjusting for their combat-exposed twin’s combat severity score no longer yielded a significant difference in handedness scores between the high-risk versus low-risk cotwins (F_(1,54)=1.3, p=0.13), although the least-square means remained in the predicted direction (47.2 for high-risk versus 53.4 for low-risk twins). The Sobel test revealed that combat severity in combat-exposed twins was a significant mediator of the association between handedness in combat-unexposed cotwins and PTSD in combat-exposed twins (z=2.1, p=0.04).

Discussion

Our results support the hypothesis that in predominantly right-handed subjects, mixed-handedness is a risk factor for combat-related PTSD and is not an acquired alteration resulting from the combat exposure or subsequent disorder. The high-risk, unexposed cotwins of Vietnam combat veterans with PTSD were more mixed-handed (i.e., less right-handed) than the low-risk, unexposed cotwins of Vietnam veterans without PTSD. Many models assume that lateral preference is determined by brain asymmetries and that dextral preference is the norm. Therefore, they look for an increased prevalence of non–right-handedness in neurologically impaired groups.¹⁹ Our results are consistent with several previous studies that found evidence that subtle impairment of the nervous system places persons at greater risk for the development of PTSD upon exposure to a traumatic event.^20,21 Such impairment might result in greater emotional reactivity to traumatic events, less successful processing of the trauma, and/or diminished ability to extinguish fear responses.

Results of the Sobel test indicated that increased combat severity in combat veterans was a significant mediator of the association between handedness in their identical cotwins and PTSD in the veterans themselves. This finding leads to the interpretation that mixed-handedness acts as a risk factor for PTSD by increasing exposure to traumatic combat events, which in turn increases the likelihood of PTSD, given that PTSD is caused by traumatic events. Combat exposure is known to be genetically influenced.²² It might be that mixed-handed soldiers, for whatever reason, are more likely to volunteer for, or be assigned to, military occupational specialties, missions, or other activities that put them in harm’s way. This might be tested in prospective studies. However, caution should be exercised with regard to this interpretation of our findings because of the consideration that subjective, retrospective reporting of combat severity may have been influenced by combat-related PTSD.^23,24 In other words, combat severity and combat-related PTSD may be confounded. Unfortunately, objective measures of combat severity are largely unavailable in Vietnam veterans.

A curious feature of these results is that although they succeeded in demonstrating a significant difference in handedness between the high-risk and low-risk cotwins, they did not succeed in demonstrating a significant difference between combat veterans themselves with versus without PTSD. Thus, the results of our twin study were more supportive of mixed-handedness as a risk factor for than as a feature of PTSD. It is possible that combat training and/or experience teaches military combatants to tend to use their hands in a similar manner (e.g., in the handling of weapons), thereby moderating initial handedness differences, although we are unaware of any data directly supporting this speculation. It is also possible that failure to find a significant difference in handedness between combat veterans with versus without PTSD in this study represented a type II error, given the large confidence limit of the observed difference. This question could be further investigated in combat veteran singletons (nontwins).