Disturbances in vergence following TBI were first described by Cross [7] and Jaensch [8], both in 1945. Over the past two decades, however, there have been several clinical studies dealing more quantitatively with visual dysfunction related primarily to mTBI. A variety of oculomotor abnormalities have been reported, including those in the vergence eye movement system, briefly reviewed next.

One of the earliest studies was by Cohen et al., who investigated vergence abnormalities in two groups of patients: 26 patients with recent-onset TBI (<3 mo postinjury) and 72 patients 3 years after TBI, of the mild variety (mTBI) in most cases [9]. Convergence dysfunction was defined by patients’ near point of convergence (NPC) amplitude. The findings were similar in the two groups: it was abnormal (i.e., significantly receded) in 38 percent of the former and in 42 percent of the latter. Ciuffreda et al. found similar results in a retrospective analysis of 160 patients with mTBI [4]. Of the patients, 56.3 percent were diagnosed with a vergence abnormality, with CI being the most common type (36.7%).

Two case series have also been reported. Berne studied three cases of young adult mTBI patients, all of whom exhibited vergence dysfunctions [10]. Each manifested a receded NPC, abnormally high exophoria at near, and decreased compensatory fusional vergence ranges at near; thus, they were classified as having CI [4]. More recently, Scheiman and Gallaway presented a case series of nine patients with mTBI [11]. Five (55%) were diagnosed with CI. The remaining four patients were diagnosed with other types of vergence abnormalities.

In a hospital-based study of 51 patients with unspecified TBI (mainly mTBI), Schlageter et al. found three vergence abnormalities present as related to the phoria: 38 percent exhibited an abnormal horizontal phoria at near, 18 percent exhibited an abnormal vertical phoria at near, and 26 percent manifested an abnormal horizontal phoria at far [12].

Lastly, Hellerstein et al. studied a range of binocular vision functions in 16 patients with mTBI [13]. These findings were compared with 16 age-matched control subjects, as well as established literature values. Significant abnormalities were found for the following parameters: near phoria, distance base-in prism break point, NPC break and recovery points, and Randot stereoacuity. Binocular suppression was noted in many: 25 percent at near distance and 37 percent at far, which is consistent with the presence of a vergence anomaly.

The purpose of the present study was to assess a wide range of static and dynamic aspects of vergence in visually symptomatic patients with mTBI.

We recruited 21 nonstrabismic patients with mTBI and near-vision symptoms (e.g., intermittent diplopia, blur) and signs (e.g., receded NPC) from the University Optometric Center at the State University of New York (SUNY) College of Optometry (Table 1). Head trauma with <30 min loss of consciousness, ≥13 score on the Glasgow Coma Scale, and <24 h posttraumatic amnesiais is defined as mTBI. [14]. It accounts for 70 to 80 percent of TBI in the United States [15-17]. Individuals ranged in age from 24 to 70 years (mean ± standard deviation: 45.7 ± 3.1 yr). There were 15 females and 6 males. All had 20/25 or better corrected visual acuity at distance and near in each eye. Only subjects who had recently completed no more than four sessions of vision rehabilitation in the past 12 months were allowed to participate. Based on our clinical experience, the effect of ≤4 sessions is minimal in these patients; however, preference was given to those with no previous history of vision rehabilitation, which was true in 16 out of the 21 subjects. Table 2 shows detailed inclusion and exclusion criteria for the group with mTBI.

We recruited 10 control subjects without mTBI from the faculty and student population at the SUNY College of Optometry. They ranged in age from 18 to 67 years (36.7 ± 5.4 yr). There was no significant difference in mean age between the control and mTBI groups (t (29) = 1.57, p = 0.13). There were 7 females and 3 males; thus, the gender ratios were similar to that of the group with mTBI. Table 2 shows detailed inclusion and exclusion criteria for the control group.

Figure 1 outlines the sequence of test procedures. The distance refractive error of each subject was fully corrected with either contact lenses or, more typically, spectacles during all testing.

---

Figure 1. Sequence of research protocol procedures. PRII = Power Refractor II. Click Image to Enlarge. View as PowerPoint Slide

We compared the mean (±1 SEM) response amplitude for convergence and divergence in the control and mTBI groups. We found no significant difference in response amplitude between the control and mTBI groups for either convergence (t(26) = 0.521, p = 0.61) or divergence (t(26) = 0.075, p = 0.94). Responses were similarly accurate in both groups.

Table 4 presents the mean values (±1 SEM) for the 14 static parameters in both the mTBI and control groups. We performed an unpaired t-test for each parameter. We found significant differences reflecting response abnormality in five of the parameters in the group with mTBI compared with the control group: NPC break (t(29) = 2.298, p = 0.03) and recovery (t(29) = 2.400, p = 0.02) values, PRV break (t(28) = 2.296, p = 0.03) and recovery (t(28) = 2.168, p = 0.04) values, and stereoacuity (t(29) = 3.229, p = 0.01). We also found five parameters that exhibited predicted directional abnormality in the group with mTBI compared with the control group [22]. These included the von Graefe near phoria test (exophoric values only), near cover test (exophoric values only), base-out PA, AP, and FD.

We performed correlations for the static and dynamic vergence parameters that were significantly different between the control and mTBI groups (Table 5). We performed these to determine whether poor performance on one parameter was also reflected in any others. For the dynamic parameters, we found 6 significant correlations in the control group and 11 in the group with mTBI. For the static parameters, we found two significant correlations in the control group and two in the group with mTBI. Some, but not all, of the significant correlations were the same in both groups.

Table 6 presents the correlations found between the various static parameters that exhibited predicted abnormal directional effects, as well as stereoacuity, for the control and mTBI groups. Both groups demonstrated a significant correlation (or a strong trend) between von Graefe near phoria and near FD, between von Graefe near phoria and near AP, and between FD and near AP. There was a significant correlation in the subjects with mTBI only between the absolute value of FD and stereoacuity. However, since the stereoacuity test did not extend below 20 sec arc, subjects with a value of 20 sec arc might actually have a lower threshold, and hence be better than indicated. With these five subjects excluded, the correlation was still significant. While there was no correlation between the absolute value of FD and PA, there were two interesting findings in the group with mTBI: some subjects (20%) exhibited either paradoxical negative PA or a large range of FDs with not much evidence of PA.

Table 7 presents the individual subjects with mTBI comparative findings for the key 13 static and dynamic parameters. They were dichotomously categorized as being either “normal” or “abnormal” using the following criterion: its value had to exceed the normal group mean by greater than ±1 SEM for that specific parameter. None of the subjects with mTBI were abnormal for either only all static or only all dynamic parameters (excluding subjects TBI-V-7, TBI-V-9, and TBI-V-17, whose dynamic data were not usable because of excessive artifacts). However, three subjects (TBI-V-2, TBI-V-11, and TBI-V-18) were abnormal for all of the static and dynamic parameters, four subjects (TBI-V-3, TBI-V-13, TBI-V-15, and TBI-V-16) were abnormal for all but one of the static and dynamic parameters, and four subjects (TBI-V-1, TBI-V-4, TBI-V-10, and TBI-V-14) were abnormal for all but two of the static and dynamic parameters. Two subjects (TBI-V-8 and TBI-V-12) were normal for all of the static parameters, whereas none were normal for all of the dynamic parameters. Of the 165 possible dynamic responses across all subjects, 18 were normal (~11%); in contrast, it was approximately 29 percent across the 84 possible static responses. The least amount of parameters found to be abnormal was six (TBI-V-12), and all were dynamic. Lastly, while none of the static parameters were abnormal in all of the subjects, five of the dynamic ones were (convergence and divergence peak velocity, convergence and divergence time constant, and divergence steady-state variability).

Table 8 presents the values for the repeated NPC measurements in the mTBI and control groups. Of the subjects with mTBI, 67 percent had an NPC break value larger than found in control subjects (14/21), and 76 percent of the subjects with mTBI had recovery values larger than found in the control subjects (16/21). Both the repeated break and recovery values were lower in 52 percent of the subjects with mTBI (11/21) compared with the control subjects. We performed a one-way repeated-measures ANOVA for the above parameters to test for significant fatigue with repeated measures (control break, control recovery, mTBI break, and mTBI recovery). We found a significant difference between the repeated NPC break values for the subjects with mTBI (F(2,20,40) = 4.73, p = 0.01), with it becoming progressively receded with repetition. There were no significant differences between the remaining three parameters upon repeated measurement (p < 0.05).

There was a significant difference between the two groups with regard to stereoacuity threshold (t(29) = 3.229, p = 0.01). It was 21.0 ± 0.5 sec arc for the control subjects and 39 ± 4 sec arc for the subjects with mTBI. Of the control subjects, 90 percent manifested a stereo-acuity of at least 20 sec arc, whereas only 33 percent of the subjects with mTBI exhibited stereoacuity of at least 20 sec arc. Randot stereoacuity using the screening plate indicated at least 250 sec arc in each subject from both test groups.

The results of the current study revealed significant differences for a range of dynamic vergence functions between the mTBI and control groups. First, and never before investigated in the population with mTBI, were the objectively based dynamic parameters of vergence, which included time constant, latency, peak velocity, and steady-state response variability, as well as the clinically based prism flipper-induced response fatigue test. All subjects with mTBI manifested decreased peak velocity and related increased time constant, as well as increased latency, for both convergence and divergence. This suggests abnormally slowed sensory processing and motor responsivity reflecting an underlying neurological control signal problem. Second, the baseline prism facility values were also significantly lower than that of the control subjects. This confirms earlier findings by Scheiman and Gallaway [11], and furthermore, it is consistent with the overall abnormally slowed dynamic parameters described earlier (e.g., time constant) [1]. However, another new result was that prism facility did not exhibit significant fatigue upon repetition, despite the frequent complaint of “visual fatigue” in the population with mTBI. This suggests that other aspects of the oculomotor system may be intimately involved (e.g., accommodation). Third, the present study also highlighted several static vergence parameters that were adversely affected by mTBI. These included a significantly receded NPC, significantly reduced fusional PRV ranges, and abnormally large phoria magnitudes, as well as significantly reduced sensory stereoacuity thresholds. These last results confirmed and extended earlier studies involving these static vergence parameters and related aspects [4,9-13]. However, given the relatively small sample size of the present pilot investigation, a definitive conclusion cannot be made. Future studies are needed using a large and perhaps more diverse range of subjects with TBI.

The new finding of increased latency (~100 ms) in the group with mTBI compared with the control group deserves mention. This delay is likely to be too long to be attributed solely to the diffuse axonal damage that occurs in the coup-contrecoup impact. It may be compounded by attentional factors, with increased temporal processing time being common in the group with mTBI [15,23], as well as a more general cognitive impairment [24]. Since none of the subjects were diagnosed with either extraocular muscle paresis or palsy using conventional clinical procedures [4,19], this delay is unlikely to be attributed to gross neural innervational and/or extraocular muscle injury-related delays [25]. However, presence of a subtle, subclinical extraocular muscle disorder may be a partial contributory factor to the overall delayed response.

Correlations within the group with mTBI for the relevant static and dynamic parameters demonstrated that in many cases (~50%), if the individual responded poorly for one parameter, they also responded poorly to several others (Tables 5-6). For example, if they responded poorly for convergence, they also responded poorly for divergence and for the parameters of peak velocity, time constant, latency, and steady-state variability. The lack of significant correlations between some of the other parameters may be expected given the predicted increased variability in their dynamic responsivity.

The individual subject comparisons (Table 7) yield some interesting trends. First, nearly all subjects (20 out of 21) were abnormal on the majority (≥7) of the 13 static and dynamic parameters. Second, performance was worse for the dynamic than the static parameters (29% vs 11%, respectively). Third, there were five dynamic but no static parameters that were abnormal across all subjects. There was no apparent relationship between vergence performance and demographic characteristics (e.g., mTBI etiology, age) (Table 1). Thus, there was no apparent reason why some subjects with mTBI performed better than others on the vergence tests. Lastly, these individual subject findings are consistent with the group correlations discussed earlier (Tables 5-6) in which we found more vergence abnormalities for the dynamic than the static parameters.

With regards to static vergence dysfunction, many earlier studies reported significant differences in the near phoria, NPC break and recovery values, and vergence ranges when the population with mTBI was compared with the control group [10-13]. Berne [10] and Scheiman and Gallaway [11] both reported significantly higher exophoria in the group with mTBI, which is consistent with the findings of the present study. In addition, the current study also found abnormally high esophoria in 3 of the 21 subjects with mTBI (15%). This latter new finding is consistent with the general notion that those with mTBI may present with general abnormal interactive-based binocular vision problems [4,26]. The receded NPC values found in the present study for both the break and recovery measures have been reported in numerous studies [10-13]. In the present study, 15 of the 21 subjects (71%) revealed significantly receded break values, and 18 of the 21 subjects (86%) demonstrated significantly receded recovery values. Lastly, there was a significant reduction in positive fusional vergence (PFV) ranges at near. This is consistent with the majority of those with mTBI having exophoria at near, thus requiring additional PFV to attain bifoveal fixation. Furthermore, it is consistent with earlier studies [10-13].

With regard to dynamic vergence dysfunction in mTBI, the results of the present study provide considerable new information. The only dynamic vergence test that has been previously performed in the population with mTBI was the baseline clinical vergence prism facility test, a subjective test. In the current study, 11 of the 21 (52%) subjects with mTBI manifested reduced vergence prism facility rates. The new objective dynamic findings of the current study can help explain the reason for the decreased clinical flipper rate; all dynamic parameters were found to be abnormal in the group with mTBI, thus resulting in slowed and delayed dynamic responsivity. These individual dynamic parameters each contributed to the overall reduced prism facility rate values. That is, the reduced clinical flipper facility rate represents an overall, more global reflection of the vergence dynamic dysfunction, whereas the results of the objectively based vergence dynamics reveal the relative degree of abnormality for each of the specific components.

With regard to stereoacuity, the results of the present study are consistent with earlier investigations [12-13]. They too found increased stereoacuity thresholds in the range found in the present study. The average stereoacuity found for the population with mTBI was 39 sec arc, with 12 out of 21 patients (60%) having 40 sec arc or worse stereoacuity, which is considered to be clinically abnormal [19]. Slightly reduced stereoacuity is consistent with the large phoria many subjects exhibited, because there is a relationship between the phoria with both the FD direction and magnitude in control subjects [27]; the greater the phoria, the greater the FD, with both being in the same direction. Our new finding was that this rela tionship was also true for the group with mTBI (Table 6).

The wide array of both dynamic and static vergence abnormalities reported in individuals with mTBI [23] may have major adverse consequences for a range of vision-related activities, such as reading, visual scanning, and tracking in depth, as well as in more general activities of daily living [4,6,22,26]. Furthermore, abnormal vergence may also interfere with performance at the workplace (i.e., have an adverse vocational effect), such as performing sustained computer-related activities, which may in return result in loss of income and related employment benefits. Moreover, it can lead to inadequate progress in other rehabilitative services (e.g., cognitive therapy) involving a range of general, as well as specific, visual eye-tracking tasks and demands [24,28].

There are several important clinical implications based on the findings of the present study. First, the objective dynamic results confirm and extend earlier clinical findings. For example, Scheiman and Gallaway [11] found reduced overall vergence facility rates, which were confirmed and extended both objectively and clinically in the present study. However, the new objective findings of the present investigation allow us to dissect and assess the individual component’s relative contribution to the overall vergence facility dysfunction. In the present study, each of the dynamic components to varying degrees were found to be abnormal: increased latency, increased time constant and related reduced peak velocity, and increased steady-state response variability. Hence, all components of fast vergence control were impaired, with them encompassing and reflecting slowed visual neurosensory processing, as well as abnormal and/or slowed and more variable motor processing and responsivity. The fact that the vergence response amplitude was appropriate and accurate for the stimulus demand suggests a relatively normal amplitude step neurological control signal [29], although the increased steady-state vergence response variability suggests increased but subtle neural noise (i.e., variability) in the step signal [30]. However, the concurrent reduced peak velocity suggests a considerable pulse-like neurological control signal deficit (i.e., reduced amplitude and/or duration), which would allow the system to eventually acquire the target, but with a slower overall dynamic time course [30-31]. Thus, both the step and pulse vergence neural controller signal components appear to be adversely affected by the mTBI.

Second, the above dynamic deficits have an important effect on one’s vision rehabilitation strategy. Namely, all aspects of fast dynamic vergence control were adversely affected, and hence should be targeted using different amounts and directions of disparity step stimuli (e.g., prism flipper step stimuli) [11]. In addition, the treatment implications extend to the static vergence domain. For example, the receded NPC and its visual fatigue with repetition, reduced and restricted fusional PRV ranges, abnormal near phoria magnitude, and increased FD magnitudes all need to be addressed clinically. They may be helped by appropriate therapeutic intervention, namely small steps of disparity stimuli, as well as smooth and continuous disparity ramp stimuli (e.g., vectograph ramp stimuli) per models of the vergence system [22,31]. Furthermore, the present abnormal static findings suggest, and are consistent with, disturbance of slow vergence control [32-34] (e.g., ramp stimulus for NPC testing). Thus, both fast and slow vergence control appear to be adversely affected by the brain injury.

Third, the abnormal static and dynamic vergence findings are consistent with the symptoms reported by these patients, such as intermittent diplopia, running together and apparent “movement” of lines of print, and vergence-induced blur, to name a few. Thus, reduction in the number of symptoms and their intensity should be correlated with improvement in clinical signs [35] and dynamic responses [36].

Lastly, the results of the present study suggest the following high yield tests: NPC with repetition, baseline prism vergence facility, near horizontal phoria, and fusional PRV ranges. These four tests are also consistent with the diagnosis of CI [19], which is commonly (56.3%) found in the population with mTBI [4]. We have recently developed such “targeted” oculomotor-based diagnostic protocols for assessment in the population with mTBI [37-39].

There were three possible limitations to the present study. First, the binocular sampling rate of the PRII was 12.5 Hz, and thus a discrete sample of dynamic eye position was obtained every 80 ms, which is sufficient for a relatively slow system such as vergence with its ≥1 s overall response time [1,40]. However, with a higher sampling rate (e.g., 25 Hz), the estimate of response latency may be slightly improved, and furthermore, the dynamic trajectory may be better resolved and quantified. However, these are likely to be second-order effects. Second, we did not perform brain imaging. This would have provided important information regarding the precise sites of damage to the brain, especially as related to vision and, more specifically, to vergence control, as mentioned previously. However, since all were mTBI, it would be expected to be relatively comprehensive in nature, thus having multiple sites of injury per the coup-countercoup aspect [23] frequently found in these patients, and not be as more localized damage as suggested by the cerebrovascular accident results in this area [32-34]. Third, because of the relatively small sample size, definitive conclusions cannot be made; thus, further related investigations are warranted.