In a recent study we analyzed ocular torsional movements during forced head tilting in normal subjects using cinematography.1 We found that as the head is tilted, the eye lags behind the head in a slow rolling movement. Periodic wheel-like ocular movements occur in the direction of the head tilt but only partially correct for the lag of the eye. This results in a partial compensatory intorsion on ipsilateral head tilt and extorsion on contralateral head tilt. Our findings of rapid extorsional movements during ipsilateral head tilt and intorsional movements on contralateal head tilt contradict classic beliefs in oculomotor physiology which hold that on ipsilateral head tilt, the extorters are inhibited and the intorters stimulated, and on contralateral head tilt the intorters are inhibited and the extorters are stimulated.2 Our finding of partial compensatory torsion at the end of forced head tilt, however, is consistent with classic teaching.
To confirm our findings, and to determine the role the oblique muscles play in the various torsional movements we observed, we studied ocular torsional movements during forced head tilting in patients with superior oblique palsy before and after inferior oblique recession.
Methods and Materials
This clinical series consists of three patients with unilateral superior oblique palsy examined and treatedby one of us (BJK). All three met the criteria of developing vertical and torsional diplopia after head trauma, with a history of having had previously normal ocular motility. The diagnosis of superior oblique palsy was made by the three step test,3 the presence of a spontaneous head tilt to the contralateral side, ameasurable excyclotropia with the double Maddox rod, and the presence of superior oblique underaction and inferior oblique overaction in the affected eye. In all three patients, preoperative filming of eye movements was done during the week prior to surgery. Postoperative filming was done between four to six months postoperatively.
Case 1 was an 18-year-old male who developed a right superior oblique palsy after a skiing accident. Preoperatively he had 12 diopters of right hypertropia in the primary position.
Case 2 was a 24-year-old male who developed a left superior oblique palsy after an automobile accident. Preoperatively he had 15 diopters of left hypertropia in the primary position.
Case 3 was a 27-year-old male who developed a left superior oblique palsy after a motorcycle accident. Preoperatively he had 18 diopters of left hypertropia in the primary position.
For the purposes of analyzing data in this study, the measurements for Case 1 were converted to represent a left superior oblique palsy. This allowed for greater ease in comparing graphs of the data.
All three patients underwent a recession of the ipsilateral inferior oblique muscle to a point 3 mm posterior and 2 mm temporal to the temporal end of the inferior rectus insertion. Postope r at i ve I y they were all considered "cured" because they had more normal versions than preoperatively, less horizontal incomitance, and elimination of their symptoms of diplopìa, Postoperatively all had less than 4 diopters of hypertropia in the primary position (range 0-4).
Our method of photographing ocular torsional movements, and our method for eliminating the artifact of false torsion, has been described in detail.1 In summary, it consists of marking the cornea with an egg membrane and photographing the eye at 24 frames per second during forced head tilt. Reference markers on the forehead and in the background allowed for calculation of rotation of the eye and rotation of the head. Frame -by- frame analysis was performed and the movement of the head, as well as rotation of the eye, were plotted against time. Our measurements were found to be accurate within one degree variance.
For consistency, the rapid wheel-like eye movement occurring at the time of greatest head velocity was used for calculating the size and velocity of this phase of ocular torsion. This was usually the greatest and fastest wheellike movement and occurred during the middle of the tilt. The slow rolling movement just prior to and just after this particular wheel-like movement was chosen for analysis.
All of the "p" values for comparison of means in this study were obtained using the t-test. For each subject the different values obtained from each of the head tilts being analyzed were averaged. This was done for each parameter being studied. This mean value for each subject was then used to represent that subject in calculating the t-test.
Our data consist of the results of three ipsilateral and three contralateral head tilts in each of three patients, preoperatively and postoperatively, totaling nine preoperative and nine postoperative head tilts.
FIGURE IA: (Kushner, Kraft, Vrabecl Frante-by-frame analysis of head movements and eye movements of normal subject during forced head tilt left. Stow rolling movements keeping the eye level with the horizon iintorsion) are seen (small arrows). Rapid wheellike mouemenis in the direction of head tii£ (estorsioni are seen (large arrows).
FIGURE iB: tKushner, Kraft, Vrabeci. The slow rolling movements which keep the eye level with the horizon iintorsion) come from the superior oblique and possibly the superior reclus Uopi. The rapid wheel-like movements in the direction of head tilt !extorsion) come from the inferior oblique and possibly the inferior reclus.
FIGURE 2A: (Kushner, Kraft, Vrabect. Frame-by-frame analysis of left eye movement of Case 1 on head tilt left. The slow rolling movements (intarsionl no longer keep the eye as perfectly steady with the horizontal as in the normal (small arrows!. The rapid wheel-like movements in the direction of head tilt (extorsion! are longer and faster than in the normal f large arrows/.
FIGURE 2B: IKushner, Kraft, Vrabect. ???? a left superior oblique palsy, the slow rolling movements fintorsion! are less perfect in holding the eye horizontal than in normal as the superior oblique, which is responsible for the movements, is weak (topi. The rapid wheel-like movements (estorsioni are greater than normal as they are caused by the overacting inferior oblique !bottom}.
The results of the analysis of ocular tors ion al movements duringforced head tilt from three normal subjects from our previously reported study' were used for comparison.
All head tilts were approximately 40° ( mean 40.43°± 1.15 ) with a mean velocity of 23.6° per second ±1.17.
For a normal subject undergoing a 40° head tilt at 23.6° per second, the eye lags behind the head at times in a slow rol ling movement keeping theeye perfectly steady with the horizon. The slow rolling movements last a mean of .186 seconds ±.025 (Figure 1). The periodic, rapid wheel-like movements in the direction of head tilt (extorsion for ipsilateral and intorsion for contralateral head tilt lhad a mean velocity of 54° per second ±6.02 and rotated the eye a mean ofll.6°±.57. The final wheel-like movements did not allow the eye to catch up with the head leaving a final compensatory intorsion for ipsilateral tilt of 10.3" ± .57.
A virtually identical picture was seen on contralateral head tilt with similar measurements. For the normal subjects studied, the slow rolling movement lasted a mean of .197 seconds ±0.20 and the rapid wheel-like movements had a mean velocity of 52° per second ± 7.01 and rotated the eye mean of 12. G±, 43.
In subjects with superior oblique palsy and inferior oblique overaction, on forced tilt to the ipsilateral side there were several striking differences as compared to the normal. The slow rolling movements (intorsion) did not keep the eye as perfectly level with the horizon during this phase of head tilt as shown by a slight rise in the graph of ocular rotation (Figure 2). In addition, the duration of these slow rolling movements was almost double those seen in the normal with a mean duration of .33 seconds ±.029 and this differed significantly from the duration of this phase in the normal (.025 > p > .011. Very little variation was seen in this phase between the three different tilts studied for each patient. The rapid rotary movement in the direction of head tilt (extorsion) was longer and faster (almost double) than in the normal, averaging 14.1° ±2.02 of rotation at a mean velocity of 94.77° per second ±6.77. Both of these values differed significantly from those of the normal (.01> p > .005). At the end of the head tilt, the final compensatory intorsion had a mean of 9.5° ±2.60 which did not differ significantly from that seen in the normal. There was, however, much greater variation for each patient between the three respective tilts studied. For Case 1 the final torsion varied between 9-11°; for Case 2, between 5-12°; and for Case 3, between 5-12111. Such variation was not seen in the normals.
On forced head tilt to the contralateral side in patients with superior oblique palsy, the eye initially lags behind the head (extorsion) (Figure 31. No rapid wheel-like movement was seen (intorsion) and the eye continued to lag behind the head. A similar picture was seen in all nine head tilts studied. At the end of head tilt, the mean extorsion was 8.8° ±2.6 which did not differ significantly from those seen in the normal. Considerable variation was also seen between the three tilts for each of the three patients. Case 1 varied between 9-12°; Case 2, between 5-10°; and Case 3 between 6-12°, for final compensatory extorsion.
FIGURE 3A: lKushner. Kraft, Vrabec). Frame-by-frante analysis of left eye movements of Case 1 an forced head tilt right. Ax the head is tilted, the eye lags behind the head !extorsion!. Rapid wheel-like movements !intarsiati! are not presen I.
FIGURE 3B; lKushner, Kraft. Vrabec). Slow rolling movements !extorsion) are present on contralateral till in patient with superior oblique palsy because the inferior nblique in functioning !bottami. Rapid wheel-like movements lintorsionl are not seen because the superior oblique is parelic ftopl.
FIGURE 4A. IKushner, Kraft, Vrabecl. Frame-by -f rame analy sis af left eye of Case 1 after inferior nblique recession on forced head tilt left. The slow rolling movements Hntorsion) are essentially unchanged from preoperatiue appearance !small arrows). The rapid wheel-like movements (extorsion/ are slower and of less magnitude than preoperatiuely !large arrows).
FIGURE 4B: lKushner, Kraft. Vrabecl- On forced head lilt left after inferior oblique recession for superior oblique palsy, the slow rolling movements I intorsioni are still weaker than in normal because the superior oblique is paretic /topi. The rapid wheel-like movements !extorsion! are slower than preoperatively in the normal, as the inferior oblique in iatrogemcaily weakened !bottom).
FIGURE SA: (Kushner. Kraft, Vrabecì. Frame-by-frame analysis of left eye movements of Case! on. farced head tilt right after left inferior oblique recession. There are multiple shorter, slow rolling movements (extorsion) than seen in the normal (small arrows). Rapid wheel-like movements /intorsionl which were absent preoperatively are now present, however, they are slower and of shorter duration than in the normal (large arrows).
Pos !Operative Iy, on ipsilateral head tilt, the slow rolling movement (intorsion) did not keep the eye perfectly steady with the horizon and tended to last longer than normal, a mean of .34 seconds ±.034. This was not significantly different from the values for the same patients p réopérât i ve Iy, but significantly differed from the values in the normal subjects (.025 > p > .01). The rapid wheel-like movements (extorsion) had a mean velocity of 36.5" per second ±3.61 which was significantly slower than both the preoperative values for the same patients and the normals (.01 > p > .005). The wheel-like movements rotated the eye a mean of 10° ±1.32 which did not differ significantly from the values of the normal subjects, but was significantly less than the values for the same patients preoperatively ( .01 > p > .005 ). The mean compensatory intorsion at the end of head tilt was 10" ±1.32. This did not differ significantly from the values seen in the normals nor in the same patients seen preoperatively (Figure 4).
The findings in the patients postoperatively on contralateral head tilt were somewhat variable (Figure 5). In general, they were characterized by more numerous, but shorter duration, wheel-like movements in the direction of head tilt (intorsion) than seen in the normal. The slow rolling movements (extorsion) in general were present, but also seemed more variable and of shorter duration. Because of the greater variability between head tilts, and the smaller size and duration of the movements being studied, we did not feel accurate mathematical analyses could be performed on these movements. In general, however, the movements of the eye postoperatively (Figure 5) more closely resembled the picture seen in the normal (Figure 1) than in the same patient preoperatively (Figure 3). In some head tilts, the eye did not return to its original orientation in the orbit after the head was straightened, leaving as much as 8° of torsional misalignment. This was never seen in normal subjects. The mean compensatory extorsion at the end of head tilt seen in these patients postoperatively was 9.5" ±2.5 which did not differ significantly from the values seen in the normal subjects or the same patients preoperatively.
FIGURE 5B: (Kushner, Kraft, Vrabec). On forced head tilt right in patient with left superior oblique palsy post left inferior oblique recession, the rapid wheel-like movements (intorsioni return but are weaker than normal (topi. The slow rolling movements (extorsion) are present but smaller and less distinct than in normal after the inferior oblique has been surgically weakened.
Classically, as the head is tilted to the ipsilateral side, compensatory intorsion occurs due to contraction of the superioroblique and the superior rectus, while the inferior oblique and inferior rectusare inhibited. The converse happens on contralateral head tilt resultingin partial compensatory extorsion. Scott1 confirmed this concept using electromyography. Ourstudy, in part, contradicts this concept. We found that during ipsilateral head tilt, the intorters (mainiy superior oblique) act to steady the eye with respect to their horizon. These movements are therefore approximately equal to the speed of head tilt, which in our study are approximately 23° per second, and more closely resemble pursuit movements than saccades. The more rapid eye movements on ipsilateral head tilt are extorsion movements mainly due to the inferior oblique. Despite the more active movement being extorsion, the eye ends up with compensatory intorsion due to the final extorsional movement being incomplete. The converse happens on contralateral head tilt. Why our data contradict those of Scott4 is not clear. Jam pel'5 felt that Scott's data showed isometric contractions, however, he did not elaborate on why he came to that conclusion. Thus, it appears that both the superior oblique and inferior oblique play a majorrole in the dynamics of torsional movements during forced head tilt to both the ipsilateral and contralateral side.5 In part, they serve to steady the eye with respect to the horizon, possibly to maintain good visual acuity. In part, they serve to keep the eye oriented approximately close to its original orientation in the orbit. However, they are incomplete in doing so.
We were surprised to see no significant difference in the amount of compensatory torsion at the end of head tilt between the normal subjects and the patients with superior oblique palsy both prior to and after surgery. The functionalimportanceofthis compensatory torsion at theend of head tilt remains obscure to us .
The Bielschowsky head tilt phenomenon*1 is classically explained on the assumption that during ipsilateral head tilt, the superior rectus and superior oblique are both stimulated. The paretic superior oblique does not balance the vertical action of the superior rectus so the eye elevates. Our data show that during ipsilateral head tilt, both the superior and inferior vertical muscles are active. Therefore, it is possible that the overacting inferior oblique, seen commonly in superior oblique palsy, is partly responsible for the elevation of the eye on ipsilateral tilt. The frequent observation that a positive head tilt test often becomes normal after ipsilateral inferior oblique recession suggests that the inferior oblique may play a role in the Bielschowsky head tilt phenomenon. Normalization of the positive Bielschowsky head tilt test should not be seen after inferior oblique recession if the inferior oblique is in fact inhibited as is classically described. In our study, we attempted to correlate the rise of the eye during forced head tilt in patients with superior palsy with the different phases of ocular torsional movement. Unfortunately, the size of the vertical movements was too small to accurately measure and note differences between successive frames of the movie.
We were also surprised to see that in some patients with superior oblique palsy, or superior oblique palsy postsurgery, the eye did not always return to its original orientation in the orbit after the tilted head was straightened. None of the patients reported any subjective symptoms in this regard,
In a series of articles, Jampel investigated the role of the oblique muscles and ocular torsional movements in monkeys and humans. s·7-9 Jampel concluded that compensatory ocular torsional movements do not occur. Our current and previously reported data,1 contradict Jampel's conclusions. Jampolsky10 has pointed out problems in using the monkey as a model for human strabismus and ocular rnotility. In addition, the retractor bulbi muscle present in monkeys, and not present in humans, may play an undetermined role in affecting ocular motility. Jampel reported7 that if the superior oblique muscle is severed surgically in a monkey, no deficit is seen in infraduction and the monkey does not assume a head tilt. Clearly, the monkey must have either a different sensory or motor system than the human, which routinely shows a hypertropia and a compensatory head tilt after superior oblique palsy. The reasons for the differences between our findings on ocular torsional movements in humans using cinematography, and those of Jampels, are not clear.
Barbara Klein, M. D. and Paul Kauf m an, M. D. assisted with the statistical analysis.
1. Kushner BJ, Kraft S: Ocular torsional movements in normal humans. Am J Qphthaimol 1983; 95:752-762.
2. Scott A: Ocular motility, in Records R (edl: Physiology i>f the Human Eye and Visual System. Hagerstown, MD, Harper and Row, 1979, pp 577-642.
3. Parks MM: Isolated cyclovertical muscle palsy. Arch Ophthalmoi. 1958; 60:1027-1035.
4. Scott AB: Extraocular muscles and head tilting. Arch Ophthalmoi 1967; 78:397-399.
5. Jampel RS; Ocular torsion and the function of the vertical extraocular muscles. Am J Ophthalmoi 1975; 79; 292-304.
6. Bielschowsky A: Lectures on motor anomalies of the eyes. Arch Ophthalmoi 1935; 13:34-59.
7. Jampel RS: The action of the superior oblique muscles. Arch Ophthalmoi 1966; 75:535-544.
8. Jampel BS; The fundamental principle of the action of the oblique ocular muscles. Am J Ophthalmoi 1970; 69:623-638.
9. Jampel RS: Ocular torsion and the primary retinal meridians. Am J Opkthalmol 1981; 91:14-24.
10. Jampolsky A: Unequal vision inputs and Strabismus Managements: A Comparison of Human and Animal Strabismus, in; Symposium on Strabismus, Transactions tif the New Orleans Academy of Ophthalmology. St Louis, CV Mosby. 1978, pp 476-4800.