ORIGINAL ARTICLE

The Clinical Course of Intermittent Exotropia

ROBERT P. RUTSTEIN, OD, MS, FAAO and DAVID A. CORLISS, OD, PhD

School of Optometry, University of Alabama at Birmingham, Birmingham, Alabama

ABSTRACT: Purpose. To report the clinical course for patients with intermittent exotropia. Methods. The clinicalrecords of patients diagnosed with intermittent exotropia from 1983 to 1991 who had at least 4 years of follow-up werereviewed. All patients with neurological or medical abnormalities, developmental delays, ocular disease, or havingstrabismus surgery during the follow-up period were excluded. Results. Of the 468 records reviewed, 73 met theinclusion criteria. Forty-four patients were female, and 29 were male. Fifty-two patients had basic intermittentexotropia, 11 patients had divergence excess intermittent exotropia, and 10 patients had convergence insufficiencyintermittent exotropia. The mean age at initial visit was 20 years (range, 1 to 63 years). The mean follow-up was 10years (range, 4 to 23 years). Four patients had amblyopia of 20/30 or worse, nine patients had a vertical deviation inthe primary position, and 10 patients had undergone extraocular muscle surgery before coming to our clinic. Sixtypatients received some form of treatment during follow-up. The mean stereoacuity at the initial and final visits were 59and 70 s arc, respectively. The initial mean spherical equivalent refraction was �0.48 D and increased to �1.15 D atthe end of the study. The mean exodeviation changed from 17.2 � at distance and 17.6 � at near at the initial visit to13.7 � at distance and 13.5 � at near at the final visit. At the initial visit, 63 patients were exotropic and 10 patientswere either heterophoric or orthophoric at distance, whereas, 60 patients were exotropic and 13 patients were eitherheterophoric or orthophoric at near. At the final visit, 37 patients were exotropic and 36 patients were eitherheterophoric or orthophoric at distance whereas 33 patients were exotropic and 39 were either heterophoric ororthophoric at near. One patient was esotropic at near at the last visit. Changes in the size and quality of theexodeviation, although statistically significant (p < 0.001), were not associated with any specific treatment regimen orwith longer periods of follow-up. Measurements exhibited a regression toward the mean. Conclusion. Intermittentexotropia improved for many patients quantitatively and qualitatively over time. That the improvement was unrelatedto any treatment and length of follow-up suggests that the changes at least quantitatively are not associated with anyphysiologic process and may be due, in part, to regression toward the mean. (Optom Vis Sci 2003;80:644–649)

Key Words: intermittent exotropia, exophoria, follow-up, progression, regression toward the mean

Intermittent exotropia is an ocular deviation that at times iscompletely controlled by positive fusional vergence and pre-sents as an exophoria and at other times is not controlled by

positive fusional vergence and presents as an exotropia. Exophoriasupposedly precedes the development of the intermittent exotro-pia, but this is probably not always the case. It is a relatively com-mon form of strabismus occurring in about 25% of all strabismiccases and in 1% of the general population.1–3 Its age of onset variesbut is often between 6 months and 4 years.4

The clinical course of intermittent exotropia is unclear. It hasbeen reported to be a progressive disorder.5 Progression or deteri-oration over time may take different forms. The intermittent exo-tropia may increase in size, may spread to another fixation distance,may become more frequent, or may become constant with loss ofbinocular vision and stereopsis. Some of the factors leading toprogression include decreased tonic convergence, decreased ac-

commodation, suppression, and increased separation of the orbitswith age.

We have noted that some patients with intermittent exotropiado not progress over time and may even improve. There are pa-tients who manifest intermittent exotropia in early childhood andare exophoric as adults. Hiles and associates6 followed 48 patientswith intermittent exotropia for an average period of 11.7 years.During the follow-up period, no patient received strabismus sur-gery. As many as 80% received some vision therapy. The averagedeviation at distance decreased from 23 to 18 �, and the averagedeviation at near decreased from 11 to 8 �. Forty patients (83%)remained within 10 � of their original distance measurements atthe end of the study. In addition, 31 of the 48 patients (64%)ceased to manifest exotropia and were heterophoric at the last visit.von Noorden7 followed 51 patients with intermittent exotropiawho were not treated for an average of 3.5 years. In 25% of the

1040-5488/03/8009-0644/0 VOL. 80, NO. 9, PP. 644–649OPTOMETRY AND VISION SCIENCECopyright © 2003 American Academy of Optometry

Optometry and Vision Science, Vol. 80, No. 9, September 2003

cases, the intermittent exotropia either remained unchanged orimproved. The purpose of this study was to report the clinicalcourse of patients with intermittent exotropia examined in ourclinic.

METHODS

After approval from our institutional review board, the clinicalrecords of all patients diagnosed with intermittent exotropia fordistance and/or near at the University of Alabama at BirminghamSchool of Optometry Binocular Vision Clinic from 1983 to 1991were reviewed. All patients had to have a follow-up for �4 years.Records of patients with neurological or medical abnormalities,developmental delays, ocular disease, or having had strabismussurgery during the follow-up period were excluded. Informationretrieved from the records included patient gender, age at initialvisit, number of visits, length of follow-up, presence of amblyopia,presence of a vertical deviation in primary position, history ofstrabismus surgery before the initial visit, whether or not othertreatment was given during follow-up, and the type of treatment,the magnitude, and type of ocular deviation at distance and near,the stereoacuity, and the refractive error at the initial and finalvisits.

For this study, amblyopia was defined as best-corrected visualacuity of 20/30 or poorer in the deviating eye. The magnitude ofthe deviation was determined with the alternate prism cover test atdistance and near utilizing accommodative targets and with thepatient wearing their refractive correction. Stereopsis was deter-mined at near using either the Titmus or Randot stereotest. Themean spherical equivalent for each patient was calculated and used

as the refractive error. All clinical measurements were taken eitherby one of the authors (RPR), optometry clinical faculty experi-enced in binocular vision anomalies, or senior optometry interns.

RESULTS

Of 468 records reviewed, the records of 73 patients, 44 femalesand 29 males, were used in the study. Table 1 characterizes subjectage distribution and follow-up. The 73 patients, ranging in agefrom 1 to 63 years at their first visits, were seen a total of 572 times.The time between the first and last recorded visits ranged from 4 to23 years, with an average of 10 years. The elapsed time betweensuccessive visits averaged about 1 year.

The type of intermittent exotropia was classified according to itsmagnitude at distance and near. Fifty-two patients (71%) had abasic exodeviation with the deviation at both distances beingwithin 10 �; 11 patients (15%) had a divergence excess exodevia-tion with the deviation at distance exceeding the deviation at nearby 10 � or more; and 10 patients (14%) had a convergence insuf-ficiency exodeviation with the deviation at near exceeding the de-viation at distance by 10 � or more.

Four patients (5%) had amblyopia of 20/30 or poorer (range,20/30 to 20/70). Three of these patients had anisometropia of 1 Dor more with the largest refractive error occurring in the deviatingeye. The other amblyopic patient had low myopic refractive errorin each eye. The low number of patients with amblyopia is consis-tent with other studies.8, 9

Nine patients (12%) manifested a vertical deviation in primaryposition in addition to the intermittent exotropia. One patient hada dissociated vertical deviation with LN. She had strabismus sur-

TABLE 1.Patient age and follow-up in years.

Minimum Maximum Mean Median

Age at first visit 1 63 20 11Age at last visit 8 72 30 22Age difference between successive visits �1 14 1 1Age difference between first and last visits 4 23 10 9Number of visits per patient 2 21 8 7

TABLE 2.Mean magnitudes of the deviation, the differences between the means of the magnitudes at the first and last visits, andthe dependent t-test results.a

Distance at First Visit (17.15 �) Distance at Last Visit (13.72 �) Near at First Visit (17.63 �)

Distance at Last Visit (13.72 �) Difference � �3.43 � — —t � 3.46p � 0.001

Near at First Visit (17.63 �) Difference � 0.49 � Difference � 3.92 � —t � �0.48 t � 3.30p � 0.633 p � 0.001

Near at Last Visit (13.46 �) Difference � �3.70 � Difference � �0.26 Difference � �4.18 �t � 2.77 t � 0.22 t � 4.11p � 0.007 p � 0.823 p � 0.001

a Differences are expressed as row minus columns df � 73 for all cells.

Clinical Course of Intermittent Exotropia—Rutstein & Corliss 645

Optometry and Vision Science, Vol. 80, No. 9, September 2003

gery at 1 year for infantile esotropia. At the first visit in our clinicwhen she was 8 years old, she manifested intermittent exotropia atdistance and near and had 80 s arc stereopsis.

Ten patients (14%) had undergone strabismus surgery beforebeing examined in our clinic. Two patients had surgery for esotro-pia and subsequently developed intermittent exotropia. Eight pa-tients had recurring intermittent exotropia.

Sixty patients (82%) had nonsurgical treatment prescribed atsome time during the follow-up period. Thirty-six patients (49%)had vision therapy, 25 (34%) were prescribed prism, and 20 (27%)were prescribed over-minus lenses. Two patients (3%) receivedocclusion therapy for amblyopia. Twenty-four patients (33%) re-ceived more than one type of treatment. Thirteen patients (18%)received no specific treatment for the intermittent exotropia dur-ing the follow-up period.

The mean stereoacuities at the first and final visits were 59 and

70 s arc. At the initial visit, 50 of the 52 patients tested had mea-surable stereopsis compared with 44 of the 46 patients tested at thefinal visit. Only one patient had no measurable stereopsis at boththe first and last visits. Another patient who had stereopsis initiallydid not have stereopsis at the last visit. For both patients, themagnitude of the exodeviation decreased by more than 10 � at onefixation distance.

The mean spherical equivalent refraction changed from �0.48to �1.15 D. At the first visit, 35 patients had a spherical equivalentrefraction ranging from �0.50 to �0.50 D compared with 20patients at the last visit. The number of patients having �1 Danisometropia either in the spherical or astigmatic components oftheir refraction was similar at both visits.

Table 2 shows the mean magnitudes of the exodeviations, thedifferences between the mean magnitudes at the first and the lastvisits, and the dependent t-test results. Overall, the intermittent

FIGURE 1.Magnitude of the distance deviation as a function of age. The linesrepresent the regression line and the 95% confidence limits for the regres-sion line.

FIGURE 2.Change in distance deviation between first and last visit vs. the timebetween the first and last visits. There is no significant relation betweenchange in magnitude and time.

FIGURE 3.Change in distance deviation between first and last visit vs. patient age atfirst visit. There is no significant relation between change in magnitudeand imputed age of onset.

FIGURE 4.Change in distance deviation between first and each successive visit vs.the time between the first and each successive visit. There is no significantrelation between change in magnitude and time within patients.

646 Clinical Course of Intermittent Exotropia—Rutstein & Corliss

Optometry and Vision Science, Vol. 80, No. 9, September 2003

exotropia became smaller, the changes being significantly different.The difference in magnitudes of the distance and near deviationsrepresent changes of 20% and 24%, respectively. The exodeviationdecreased by 5 � or more for 28 patients at distance and for 33patients at near. It stayed the same or decreased by �5 � for 21 and26 patients at distance and near, respectively. For the remainingpatients, the deviation increased. However, the increase was largerthan 5 � at either distance or near for only six patients.

The effect of treatment regimen and length of follow-up on thechanges in the exodeviation were analyzed. As shown by repeated-measures analysis of variance, there was no significant difference (F� 0.0002, p � 0.99) in the magnitude of the changes in theexodeviation at the first and final visits for patients who weretreated at some point during the follow-up period (N � 60) vs.patients who were not treated (N � 13). This could be the result ofpoor patient compliance because the analysis is based on whethertreatment was prescribed, not on the type of treatment or theamount of treatment actually done.

For the 10 patients who had strabismus surgery before beingevaluated in our clinic, the effect of surgical history on the changesin the exodeviation over time was also analyzed. As shown byrepeated-measures analysis of variance, surgical history had no ef-

fect on the magnitude of change in the distance deviation (F �0.523, p � 0.472) and the near deviation (F � 0.825, p � 0.367).

Fig. 1 shows that the exodeviation at distance had a tendency todecrease with increasing patient age. Despite the high degree ofscatter, the relation is significant (r � �0.168, p � 0.001), al-though age explains little of the variance.

Fig. 2 illustrates the change in magnitude between the first andlast visits as a function of time between those visits. There is nosignificant relation between these two variables. Fig. 3 shows thatthe change in the magnitude of the deviation between the first andlast visit is also independent of the age at which the patient was firstevaluated in our clinic. In addition, Fig. 4 illustrates the change inmagnitude of the distance deviation at each visit from the magni-tude at the first visit vs. the time between the first and each succes-sive visit. Again, there is no significant relation.

The observed differences in the magnitude of the exodeviationare possibly due, at least in part, to a measurement effect. In par-ticular, where there are functions in any kind of physiologic mea-surement due either to physiologic variability or measurement er-ror, there may exist the phenomenon of regression toward themean.10 This means that if at one visit the magnitude is higher thanthe mean, then at the next visit there is good likelihood that themagnitude will be smaller than the first measure. Conversely, if theinitial measurement is low at one visit, it may be higher at the next.

Fig. 5 demonstrates the existence of this measurement phenom-enon in our data. Fig. 5 shows the probability of an increase, adecrease, and no change in the magnitude of the exodeviation fromone visit to the next as a function of the magnitude of the deviationat a particular visit. It is clear that as the magnitude of the exode-viation increases, the probability of measuring a smaller value at thenext visit increases, whereas the probabilities of measuring asmaller value or no change decreases. The opposite is true at lowvalues of the deviation.

Concomitant with but independent of the magnitude of thedeviation, the type of deviation is classified as either an intermit-tent exotropia or phoria. Initially, 63 patients (86%) were exo-tropic and 10 patients (14%) were either heterophoric or or-thophoric at distance, whereas 60 patients (82%) were exotropicand 13 (18%) were either heterophoric or orthophoric at near.Fifty patients (68%) manifested intermittent exotropia at bothdistance and near. At the last visit, 37 patients (51%) were exo-tropic and 36 (49%) were either heterophoric or orthophoric atdistance, whereas 33 (45%) were exotropic and 39 (53%) wereeither heterophoric or orthophoric at near. Twenty-eight patients(38%) manifested intermittent exotropia at both distance and

TABLE 3.Changes in the classification of the exodeviation between the first and the last visits for both the distance andnear deviations.a

Distance Near

Exotropia Exophoria Orthophoria or Esophoria Exotropia Exophoria Orthophoria or Esophoriab

First Visit 63 (86%) 4 (6%) 6 (8%) 60 (82%) 10 (14%) 3 (4%)Last Visit 37 (51%) 26 (35%) 10 (14%) 33 (45%) 32 (44%) 7 (10%)

a All changes in deviation types are statistically significant.b One patient was esotropic at near for the last visit.

FIGURE 5.Probability of an increase, decrease, or no change in the magnitude of thedistance deviation at a subsequent visit as a function of the magnitude ofthe deviation at the current visit. Note that the probability of observing adecrease in magnitude at the next visit increases as the magnitude of thedeviation of the current visit increases.

Clinical Course of Intermittent Exotropia—Rutstein & Corliss 647

Optometry and Vision Science, Vol. 80, No. 9, September 2003

near. One patient manifested intermittent esotropia at near duringthe last visit and was prescribed bifocals.

Table 3 shows cross tabulations of the exodeviation types at thefirst and last visits. All the changes are statistically significant (Mc-Nemar �2 � 8.4, p � 0.001 for distance, and McNemar �2 �14.6, p � 0.001 for near). Fig. 6 shows the mean magnitudes of thedistance deviations associated with these quality changes. There areno changes from exophoria to intermittent exotropia. Exotropias�20 � had a higher incidence of converting to exophoria than didexotropias exceeding 20 �.

DISCUSSION

The results of this retrospective study indicate that intermittentexotropia does not progress with advancing age for many patientsand may even improve. For our patients, the first and last stereopsismeasurements were similar; the magnitude of the deviation be-came smaller by an average of 3.5 � at distance and 4.1 � at near,and for 36% of the patients, the intermittent exotropia convertedeither to exophoria or orthophoria at the last visit.

These findings are in agreement with those of Hiles and associ-ates6 and Fournier and associates11 (Table 4). Unlike the earlierretrospective studies, the present study was not limited to childrenand also included patients who had undergone strabismus surgerybefore being examined in our facility. In addition, most of our

patients (71%) had a basic intermittent exotropia, whereas mostpatients in the earlier studies had a divergence excess intermittentexotropia.

If there is a physiologic process that caused the overall improve-ment, one might expect that it would be affected by treatmentregimen and/or length of follow-up. The intermittent exotropiawould be more likely to have become smaller for patients who weretreated during follow-up vs. those who were not treated. Similarly,greater decreases in the exodeviation would be more likely forpatients having longer periods of follow-up vs. shorter periods offollow-up. We could, however, find no time dependency that onewould expect of a physiologic process to explain the lack of pro-gression and overall improvement of intermittent exotropia. Fur-thermore, the reduction in the magnitude was not related towhether patients were prescribed treatment such as vision therapy,prisms, or over-minus lenses during the follow-up period.

Fournier and associates11 treated 35 of their 65 patients witheither vision therapy, prisms, or over-minus lenses and reportedthat the average distance exotropia decreased from 21 � to 15.6 �,whereas for the 30 patients not receiving treatment, the magnituderemained relatively stable. Hiles and associates6 also prescribedvision therapy for approximately 80% of their patients. Sixty pa-tients in the present study were prescribed nonsurgical treatmentsuch as vision therapy, over-minus lenses, and/or prism at some

FIGURE 6.Mean distance deviations at first and last visits associated with deviationtype at these visits. XT, intermittent exotropia; XP, exophoria; OR,orthophoria.

FIGURE 7.Probability of an increase, decrease, or no change in the magnitude of thedistance deviation at a subsequent visit as a function of the magnitude ofthe deviation at the current visit (from Hiles and associates6).

TABLE 4.Comparison of studies on course of intermittent exotropia.

StudyNo. of

PatientsAverage Age

(yr)Average Follow-Up

(yr)D1

a (�) D2 (�) N1 (�) N2 (�)

Present Study (2003) 73 20 10 17.2 13.7 17.6 13.5Hiles et al. (1968)6 48 4.8 11.7 23 18 11 8Fournier et al. (2000)11 65 4.1 5.3 21.3 (21.0)b 20.2 (15.6)b 16.1 18.1

a D1, magnitude at distance for first visit; D2, magnitude at distance for last visit; N1, magnitude at near for first visit; N2, magnitudeat near for last visit.

b Magnitudes for 35 patients who received treatment during follow-up.

648 Clinical Course of Intermittent Exotropia—Rutstein & Corliss

Optometry and Vision Science, Vol. 80, No. 9, September 2003

time during the follow-up period. Our analysis was based onwhether treatment was prescribed as determined by record reviewrather than the actual amount of therapy that was actually done. Aprospective study that carefully monitors the amount of therapyundertaken is needed to determine the efficacy of nonsurgicaltreatment in reducing the magnitude of intermittent exotropia.

Although the magnitude of the intermittent exotropia variedsomewhat with the age of the patient, its overall reduction wasunrelated to the duration of the follow-up. In part, these changescan be attributed to a statistical phenomenon known as regressiontoward the mean.10 If at one visit the magnitude is higher than themean, than at the next visit it is likely that the magnitude will besmaller than the mean. Conversely, if the first measurement is low,it is likely that it will be higher at the next visit. This is illustratedfor our patients in Fig. 5. Regression toward the mean is notlimited to measuring ocular deviations and has also been reportedin studies monitoring intraocular pressure.12

The findings of Hiles and associates6 were digitized and replot-ted to determine whether regression toward the mean was also afactor in their study. As illustrated in Fig. 7, it is evident that as themagnitude of the deviation at the first visit increased for theirpatients, the probability of a decrease at the last visit also increased.It should be noted that the data of Hiles and associates6 involvedanalysis comparing only the first and last patient visits, whereas ourdata analyzed the magnitude of the exodeviation for each succes-sive patient visit.

Over one-third of our patients no longer manifested intermit-tent exotropia and were either heterophoric or orthophoric at thelast visit. For the 48 patients reported by Hiles and associates,6

64% were heterophoric at the last visit. However, Chia and asso-ciates13 recently reported improved control of intermittent exotro-pia in only 18% of their patients at distance and in only 6% of theirpatients at near. The magnitude of the exotropia for their patientswas generally larger and averaged 36 � at distance and 26 � at near.Combining the findings of the present study with those of Hilesand associates,6 it appears that as many as half of patients withintermittent exotropia may achieve better control over time for atleast one fixation distance. This appears more likely when theexodeviation is �20 � (Fig. 6).

In summary, intermittent exotropia improved both quantita-tively and qualitatively for many of our patients. The quantitativechanges were possibly due, in part, to the effect of regression to-ward the mean. Intermittent exotropia may not be a progressivedisorder for many patients. A long-term prospective study thatcarefully monitors the magnitude and quality of intermittent exo-tropia is needed to add to our knowledge of this type of strabismus.

ACKNOWLEDGMENTS

Presented, in part, at the annual meeting of the American Academy of Op-tometry, San Diego, California, December 2002.

Received October 17, 2002; revision received May 26, 2003.

REFERENCES

1. Cooper J, Medow N. Intermittent exotropia: basic and divergenceexcess type. Binoc Vis Eye Muscle Surg 1993;8:185–222.

2. Chew E, Remaley NA, Tamboli A, Zhao J, Podgor MJ, Klebanoff M.Risk factors for esotropia and exotropia. Arch Ophthalmol 1994;112:1349–55.

3. Newman J, Mazow ML. Intermittent exotropia: is surgery necessary?Ophthalmic Surg 1981;12:199–202.

4. Rutstein RP, Daum KM. Anomalies of binocular vision: diagnosisand management. St. Louis, MO: Mosby, 1998.

5. Jampolsky A. Differential diagnostic characteristics of intermittentexotropia and true exophoria. Am Orthopt J 1954;4:48–55.

6. Hiles DA, Davies GT, Costenbader FD. Long-term observations onunoperated intermittent exotropia. Arch Ophthalmol 1968;80:436–42.

7. von Noorden, GK. Binocular vision and ocular motility: theory andmanagement of strabismus, 5th ed. St. Louis, MO: Mosby, 1996.

8. Smith K, Kaben TJ, Orton R. Incidence of amblyopia in intermittentexotropia. Am Orthopt J 1995;45:90–6.

9. Dunlap EA, Gaffney RB. Surgical management of intermittent exo-tropia. Am Orthopt J 1963;13:20–33.

10. Streiner DL. Regression toward the mean: its etiology, diagnosis, andtreatment. Can J Psychiatry 2001;46:72–6.

11. Fournier AV, Keech RV, Havertape SA. The long-term course ofunoperated intermittent exotropia. Paper presentation at the annualmeeting of the American Academy of Pediatric Ophthalmology andStrabismus, 2000.

12. Munger R, Dohadwala AA, Hodge WG, Jackson WB, Mintsioulis G,Damji KF. Changes in measured intraocular pressure after hyperopicphotorefractive keratectomy. J Cataract Refract Surg 2001;27:1254–62.

13. Chia AW, Quah BL, Linley S. Progression of intermittent exotropiaamongst children in Singapore. Poster presentation at the annualmeeting of the American Academy of Pediatric Ophthalmology andStrabismus, 2000.

Robert P. RutsteinCollege of Optometry

University of Alabama at Birmingham1716 University Blvd.

Birmingham, AL 35294-0010e-mail: rrutstein@icare.opt.uab.edu

Clinical Course of Intermittent Exotropia—Rutstein & Corliss 649

Optometry and Vision Science, Vol. 80, No. 9, September 2003