The Course of Intermittent Exotropia in a

Population-Based Cohort
Kevin J. Nusz, MD,1 Brian G. Mohney, MD,1 Nancy N. Diehl, BS2

Purpose: To evaluate the change in the angle of deviation in an incidence cohort of pediatric patients
diagnosed with intermittent exotropia during a 20-year period.
Design: Retrospective, population-based observational study.
Participants: All pediatric (⬍19 years old) residents of Olmsted County, Minnesota diagnosed with inter-
mittent exotropia (ⱖ10 prism diopters) from January 1, 1975 through December 31, 1994.
Methods: The medical records of all potential patients identified by the resources of the Rochester Epide-
miology Project were reviewed.
Main Outcome Measures: The change in the angle of deviation and its association with treatment were
reviewed for each patient.
Results: A total of 184 pediatric patients were diagnosed during the study period, of which 138 patients
(75.0%) had ⱖ2 examinations. The deviation resolved in 5 of the 138 patients (3.6%) during a median follow-up
of 9.2 years, while the Kaplan–Meier rate of increasing by 10 or more prism diopters (PD) was 23.1% at 5 years
and 52.8% at 20 years. The distance deviation increased by a median of 5 PD during the preoperative period in
the 55 patients who underwent surgery during a mean follow-up of 3.2 years compared with a zero PD median
change in the 83 patients who avoided surgery during a mean follow-up of 7.1 years. The Kaplan–Meier
probability of undergoing surgery within 20 years after diagnosis was 74.0% in this population. We were unable
to detect a significant association between nonsurgical treatments and a change in the angle of deviation.
Conclusions: In this population-based cohort of pediatric patients with intermittent exotropia, the deviation
resolved in 4%, and more than half of the patients were expected to have an increase of 10 or more PD within
20 years of their diagnosis. Children who received surgery in this population were significantly more likely to have
demonstrated an increase in their deviation during the preoperative period. Ophthalmology 2006;113:
1154 –1158 © 2006 by the American Academy of Ophthalmology.

Intermittent exotropia is the most common form of child-
hood exotropia1–3 and is more prevalent than esotropia in
some populations.3 A recent population-based study in the
United States found that intermittent exotropia occurs in
approximately 1 in 185 children by 10 years of age.2 Al-
though common, the natural history of this disorder remains
obscure. Currently, there are no prospective studies and
only a few retrospective studies of untreated intermittent
exotropia.4 – 8 However, these reports were comprised of
selected populations, many of whom had small or medically
treated deviations.
The primary objective of this study was to evaluate the
natural history of intermittent exotropia in a population-
based incidence cohort of patients younger than 19 years of
Originally received: August 6, 2005.
Accepted: January 13, 2006. Manuscript no. 2005-736.
1
Department of Ophthalmology, Mayo Clinic and Mayo Foundation,
Rochester, Minnesota.
2
Division of Biostatistics, Mayo Clinic and Mayo Foundation, Rochester,
Minnesota.
This study was supported in part by an unrestricted grant from Research to
Prevent Blindness, Inc., New York, New York.
Correspondence to Brian G. Mohney, MD, Mayo Clinic, Department of
Ophthalmology, 200 First Street SW, Rochester, MN 55905. E-mail:
mohney@mayo.edu.
age. The change in the angle of deviation, recommended
treatment modalities, and parental observations about their
child’s deviation were evaluated in this 20-year incidence
cohort.
Subjects and Methods
The medical records of all patients younger than 19 years of age
who were residing in Olmsted County, Minnesota when diagnosed
by an ophthalmologist as having intermittent exotropia between
January 1, 1975 and December 31, 1994 were retrospectively
reviewed. Institutional review board approval was obtained for this
study. Potential cases of intermittent exotropia were identified
using the resources of the Rochester Epidemiology Project, a
medical record linkage system designed to capture data on any
patient–physician encounter in Olmsted County, Minnesota.9 The
racial distribution of Olmsted County residents in 1990 was 95.7%
Caucasian, 3.0% Asian-American, 0.7% African-American, and
0.3% each for Native-American and other. The population of this
county (106 470 in 1990) is relatively isolated from other urban
areas, and virtually all medical care is provided to residents by the
Mayo Clinic or the Olmsted Medical Group. Unaffiliated area
optometrists were rarely the sole provider of eye care to children
with strabismus during the study years of this investigation.2
Patients not residing in Olmsted County at the time of their
diagnosis were excluded. Intermittent exotropia was defined in this
study as an intermittent distance exodeviation of at least 10 prism

1154 © 2006 by the American Academy of Ophthalmology ISSN 0161-6420/06/$–see front matter
Published by Elsevier Inc. doi:10.1016/j.ophtha.2006.01.033
 Nusz et al 䡠 Course of Intermittent Exotropia

Table 1. Initial Historical and Clinical Characteristics of 184 100

Patients (⬍19 Years of Age) with Intermittent Exotropia from 95% Confidence
Interval

Olmsted County, Minnesota

P 80
Characteristics Results e
r
c
Gender e 60
Male (%) 65 (35.3%) n
Female (%) 119 (64.7%) t
Family history of strabismus (%) 61 (33.2%)
Premature birth (%)* 10 (5.4%) 40
Median birth weight in grams (range) 3374 (953–4620)
History of diplopia (%) 8 (4.3%)
Amblyopia (%) 4 (2.2%) 20
Mean age at diagnosis in yrs (range) 6.3 (0.7–18.6)
Median age at diagnosis in yrs (range) 5.3 (0.7–18.6)
Median angle of deviation at distance in prism 20.0 (10.0–45.0) 0
diopters (range)
0 5 10 15 20
Median angle of deviation at near in prism 14.0 (0.0–45.0) Years
diopters (range)
Median refraction (range) ⫹0.50 (⫺8.00 to ⫹3.10) Figure 1. Kaplan–Meier plot of percentage of patients who had an in-
crease of 10 or more prism diopters from date of diagnosis.
*Born ⬍37 weeks gestational age.
ment of 0 to 8 PD measured with alternate prism cover testing at
the most recent examination.10
diopters (PD) without an underlying or associated neurologic, Comparisons between groups for continuous variables were
paralytic, or ocular sensory disorder. completed using Wilcoxon rank-sum tests, whereas categorical
Data abstracted from the medical records included gender, variables were determined by the Fisher exact test. All statistical
family history of strabismus, birth weight, gestational age at birth, tests were 2 sided, with the threshold of significance set at P ⫽
reported age at onset, ocular findings, recommended management, 0.05. The cumulative probability of undergoing surgery, sponta-
and parental observations of progression. The angle of deviation neous resolution, or the potential for the distance deviation to
was primarily determined by the prism and alternate cover tech- increase by 10 PD or more was estimated using the Kaplan–Meier
nique at both distance and near, although some patients were method.11 Potential risk factors for these end points were evaluated
measured by the Hirschberg or modified Krimsky techniques at using Cox proportional hazard models.
near. The refractive error was determined in the majority of pa-
tients after the topical administration of 1% cyclopentolate. Sen-
sory testing, when performed, was primarily done by using the
Titmus fly and Lang stereotest. There were no specific criteria for Results
enacting medical or surgical therapy; instead, treatment was at the
discretion of the provider with parental consent. Follow-up was During the 20-year study period, 184 new cases of childhood
measured from the initial to the final preoperative examination in intermittent exotropia were diagnosed at a mean age of 6.3 years
those who underwent surgery, and from the initial to the most (range, 8.4 months to 18.6 years) in Olmsted County, Minnesota.
recent examination in those who did not. Spontaneous resolution Pertinent historical and clinical characteristics of the study patients
of intermittent exotropia was defined in this study as a misalign- are shown in Table 1. Two thirds of the patients were female,

Table 2. Comparison of Clinical Characteristics of Patients with Intermittent Exotropia Whose Deviation Resolved with Those
Whose Did Not

Patient Whose Intermittent

Exotropia Spontaneously Patients Whose Intermittent
Resolved Exotropia Did Not Resolve Hazard Ratio
Characteristics (n ⴝ 5) (n ⴝ 133) (95% Confidence Interval) P Value
Mean age at diagnosis in yrs (range) 4.8 (2.3–7.7) 5.5 (0.7–18.6) 0.87 (0.64–1.18) 0.371
Medial initial angle of distance deviation in prism 14.0 (10.0–30.0) 20.0 (10.0–40.0) 0.90 (0.76–1.06) 0.188
diopters (range)
Median final angle of distance deviation in prism 6.0 (0.0–8.0) 25.0 (10.0–50.0) 0.96 (0.85–1.09) 0.551
diopters (range)
Median change in angle of distance deviation in ⫺8.0 (⫺30.0 to 2.0) 0.0 (⫺18.0 to ⫹36.0) — —
prism diopters* (range)
Mean length of follow-up in yrs (range) 9.2 (5.4–18.6) 5.4 (0.04–25.7) — —
Number who underwent medical treatment 1 (20.0%) 28 (21.1%) — —
Abnormal versions at initial examination 1 (20.0%) 25 (18.8%) 0.75 (0.08–6.78) 0.798
Abnormal versions at final examination 0 (0.0%) 32 (24.1%) 1.01 (0.11–9.3) 0.997
Median initial stereopsis in seconds of arc (range) 65 (40–100) 80 (40–3000) 1.0 (0.98–1.01) 0.583
Median final stereopsis in seconds of arc (range) 40 (40–100) 40 (40–3000) 1.00 (0.99–1.01) 0.691

*Indicates a decrease, whereas ⫹ indicates an increase.

1155
 Ophthalmology Volume 113, Number 7, July 2006

100 Table 4. Types of Medical Management among 29 Patients

95% Confidence
Interval with Intermittent Exotropia

80 Types of Medical Management Numbers Percentages

P
e Spectacles for strabismus 1 2.8
r Orthoptics/exercises 4 11.1
c 60 Overminus spectacles 4 11.1
e
n
Spectacles with prism 1 2.8
t Alternated eye patching 3 8.3
40 Patching dominant eye 23 63.9

20
0 years), there were no significant clinical differences between the 2
0 5 10 15
Years
Figure 2. Kaplan–Meier plot of percentage of patients undergoing surgery
from date of diagnosis.
whereas 1 in 3 reported a family history of strabismus. A history
of prematurity or diplopia was uncommon, and the ocular findings
were otherwise unremarkable for this disorder (Table 1).
Forty-six of the 184 study patients (25.0%) had only one
examination and were eliminated from further analysis. The age at
diagnosis was significantly older for the 46 patients with only 1
visit (mean, 8.8 years) compared with those with ⱖ2 visits (mean,
5.5 years). Otherwise, gender, birth weight, prevalence of ambly-
opia, distance angle of deviation, refractive error, and mean age at
first surgery were no different between the 2 groups.
The remaining 138 study patients who had ⱖ2 measurements
of the exotropic angle were followed-up for a mean duration of 5.6
years (range, 13 days to 25.7 years). The median change in the
distance angle of deviation was 0 PD (range, decrease of 30 PD to
an increase of 36 PD), whereas the near angle increased by 4 PD
(range, decrease of 40 PD to an increase of 40 PD) during the
follow-up period. The Kaplan–Meier rate of increasing by 10 or
more PD was 23.1% at 5 years, 29.2% at 10 years, and 52.8% at
20 years (Fig 1). Five of the 138 patients (3.6%) experienced a
spontaneous resolution during a mean follow-up of 9.2 years.
Table 2 examines the characteristics of patients who spontaneously
resolved. Although the patients who spontaneously resolved
tended to have a smaller initial angle of deviation (14 PD vs. 20
PD) and a longer mean period of follow-up (9.2 years vs. 5.4
20 groups (Table 2). The Kaplan–Meier rate of spontaneous resolu-
tion was 0.0% at 5 years, 8.5% at 10 years, and 18.7% at 20 years.
Fifty-five of 138 patients (40.0%) underwent strabismus sur-
gery at a mean age of 7.6 years (range, 3.3 to 22.8 years). The
Kaplan–Meier probability of undergoing surgery during the study
period in this population was 43.4% at 5 years, 48.4% at 10 years,
and 74.0% at 20 years (Fig 2). Table 3 compares the clinical
characteristics of patients who underwent strabismus surgery with
those who did not. Although patients who had strabismus surgery
had a similar initial distance deviation compared with those who
did not have surgery, the age at diagnosis was significantly
younger (P ⫽ 0.008) for the children who underwent surgery
(Table 3). During the preoperative follow-up, the distance devia-
tion increased by a median of 5 PD in those who underwent
surgery, whereas the median deviation of those who did not
undergo surgery was unchanged.
Twenty-nine of 138 patients (21.0%) underwent some form
of medical therapy with dominant eye patching being the most
frequent management (Table 4). Children from this cohort who
underwent medical treatment were significantly younger in age
(P ⫽ 0.005) and had a greater initial distance angle of exode-
viation (P ⫽ 0.0003) than those who did not receive medical
therapy (Table 5). Table 6 shows the number of patients who
underwent surgery within 6 months of the parents’ reporting
that their child’s deviation was worse, the same, or better. Of
those children whose parents’ reported a worsening of their
condition, 81.8% had surgery within 6 months of the recorded
observation.

Table 3. Comparison of Clinical Characteristics of Patients with Intermittent Exotropia Who Underwent Surgery with Those Who
Did Not

Patients Who Had Patients without

Strabismus Surgery Strabismus Surgery Hazard Ratio
Characteristics (n ⴝ 55) (n ⴝ 78) (95% Confidence Interval) P Value
Mean age at diagnosis in yrs (range) 4.3 (1.0–18.6) 6.4 (0.7–14.7) 0.88 (0.80–0.97) 0.008
Median initial angle of distance deviation in 20.0 (10.0–40.0) 20.0 (10.0–35.0) 1.05 (1.01–1.08) 0.010
prism diopters (range)
Median final angle of distance deviation in prism 30.0 (10.0–40.0) 20.0 (4.0–50.0) 1.11 (1.07–1.15) ⬍0.001
diopters (range)
Median change in angle of distance deviation in ⫹5.0 (⫺10.0 to ⫹30.0) 0.0 (⫺18.0 to ⫹36.0) — —
prism diopters* (range)
Mean length of follow-up in yrs (range) 3.2 (0.1–19.0) 7.0 (0.04–25.7) — —
Number who underwent medical treatment 16 (29.1%) 12 (15.4%) — —
Abnormal versions at initial examination 12 (21.8%) 13 (16.7%) 1.11 (0.61–2.04) 0.734
Abnormal versions at final examination 20 (36.4%) 12 (15.4%) — —
Median initial stereopsis in seconds of arc (range) 100 (40–3000) 55 (40–3000) 1.00 (1.00–1.00) 0.045
Median final stereopsis in seconds of arc (range) 55 (40–3000) 40 (40–3000) 1.00 (0.99–1.00) 0.706

*Indicates a decrease, whereas ⫹ indicates an increase.

1156
 Nusz et al 䡠 Course of Intermittent Exotropia

Table 5. Clinical Characteristics of Patients Who Received Medical Management

Patients Who Patients Who Did Not Hazard Ratio

Received Medical Receive Medical (95% Confidence Interval)
Characteristics Treatment (n ⴝ 29) Treatment (n ⴝ 109) P Value
Mean age (yrs) at diagnosis (range) 3.9 (0.7–12.6) 5.9 (0.9–18.6) 0.81 (0.70 – 0.94)
P value ⫽ 0.0054
Median initial horizontal angle of distance deviation 26.5 (10.0–40.0) 20.0 (10.0–40.0) 1.10 (1.05–1.16)
in prism diopters (range) P value ⫽ 0.0003
Median final horizontal angle of distance deviation 25.0 (0.0–40.0) 20.0 (4.0–50.0) 1.07 (1.02–1.13)
in prism diopters (range) P value ⫽ 0.003
Median change in horizontal angle of distance 0.0 (⫺30.0 to ⫹15.0) 0.0 (⫺18.0 to ⫹36.0) —
deviation in prism diopters* (range)
Mean length of follow-up in yrs (range) 4.6 (0.3–18.4) 5.8 (0.04–25.7) —
Spontaneous resolution rate 1 (3.5%) 4 (3.6%) —

*Indicates a decrease, whereas ⫹ indicates an increase.

Discussion duration of follow-up. Chia et al,8 with only one third of

The findings from this cohort of children with intermittent
exotropia from Olmsted County, Minnesota suggest that the
distance exodeviation of approximately one half of patients
would increase by 10 or more PD after 20 years of follow-
up. During a median follow-up of 5.6 years, the exotropia
resolved in 3.6%, and children who received surgery in this
cohort were significantly more likely to demonstrate an
increase in their preoperative angle of deviation than those
who did not undergo surgery. Nearly 3 of 4 children were
expected to undergo surgical correction within 20 years of
diagnosis. We were unable to determine any significant
association between the use of medical therapy and a
change in the angle of deviation.
Although intermittent exotropia is a common disorder,
currently there are few data available regarding its natural
history. Most studies, none of which are population-based,
are plagued by selection bias and variable degrees of non-
surgical management. Perhaps the most frequently cited
article is that by Hiles et al.4 They reported a mean distance
exodeviation decrease of 5 PD in 48 unoperated patients
followed-up for an average duration of 11.7 years. How-
ever, these patients had relatively small or well-controlled
deviations. Moreover, 81% of the patients received some
form of medical therapy.12 Although the significance of this
form of intervention is unknown, orthoptic therapy for exo-
tropic patients with deviations less than 30 PD has been
reported to attain a level of long-term success comparable
with surgery.13–15 In contrast, only 21% of the patients in
this study were managed with medical intervention, some of
whom later required surgical correction. More recently,
Rutstein and Corliss7 have reported similar findings (3 PD
mean decrease in the distance exodeviation in 73 patients
followed-up for an average of 10 years). However, a similar
proportion of their patients (82%) received medical therapy,
and because the authors do not perform surgery, their pa-
tient population may have been biased toward those who
were less likely to require surgical intervention.
Other studies, also comprised of selected populations,
have reported a moderate increase in the magnitude of the
deviating angle, although the results vary according to the
their 287 patients having 3 or more years of follow-up,
found 47.2% remained within 5 PD of their initial exode-
viation, and 33.7% deteriorated by more than 5 PD. In a
report by von Noorden and Campos,6 75% of 51 untreated
patients with intermittent exotropia (age range, 5 to 10 years
old) demonstrated an increased magnitude of deviation
and/or progressive loss of binocular function for an average
follow-up of 3.5 years. No change occurred in 9% of the
patients, and 16% improved without treatment.6 Kii and
Nakagawa5 have observed a larger angle of exodeviation in
their preoperative adult patients compared with their pedi-
atric patients, thereby supporting the notion that the strabis-
mic angle increases with time in intermittent exotropia.
However, they noted no significant change in the angle of
deviation among children followed-up for a 5-year time
interval.5
In our population-based cohort of children with intermit-
tent exotropia, approximately one half of the patients were
calculated to have an increase of 10 or more PD after 20
years of follow-up. Moreover, during the same 20-year
interval, nearly three quarters of patients were expected to
undergo surgery based on the practices of our department
during the years of this study. Atlhough the study surgeons
did not specifically postpone surgery until a prescribed
increase in the deviation was observed, the finding that half
of the study patients were calculated to increase by 10 or
more PD within 20 years may have influenced the examin-
ers toward recommending surgery.
We found a relatively low rate of spontaneous resolution
in this population of children with intermittent exotropia.
Approximately 4% of patients resolved during a median
follow-up of 9 years, compared with the recent findings for
Table 6. Parental Observations Recorded in the Patient History
within 6 Months of Strabismus Surgery
Parental Observation Surgery within 6 Months of
of Child’s Exotropia Parental Observation (n ⴝ 33)
Worse 27 (81.8%)
Same or better 6 (18.2%)

1157
 Ophthalmology Volume 113, Number 7, July 2006
congenital esotropia in which 27% of 170 children experi-
enced a resolution within months of their diagnosis.10 We
predict a higher rate of spontaneous resolution for patients
with intermittent exotropia given that the median follow-up
for all patients in this study was significantly less than for
the 5 patients whose deviation resolved.
Only a small number of patients in this study were
managed with nonsurgical therapy. Although there was no
significant difference in progression between patients who
received medical therapy and those who did not, it is diffi-
cult to arrive at any definitive conclusions based on the
small numbers who were treated. We were also unable to
make any conclusions regarding the influence of parental
observations on the timing of surgery, although there may
be an association between parental input and the subsequent
management recommendation.
There are a number of limitations to the findings of this
study. Patients undergoing surgery were eliminated from
further follow-up, potentially biasing the data toward those
with smaller angles of deviation, presuming that those with
larger angles underwent surgery. However, a purely obser-
vational study, devoid of any form of management, would
be potentially harmful to at least some of the study patients
should they become constant and lose stereopsis during the
study period. Even the vast majority of the unoperated
subjects by Hiles et al4 received some form of medical
therapy.12 Moreover, the Kaplan–Meier calculations used in
this study have been shown to assess longevity accurately in
subjects with unequal follow-up.11 This report is also lim-
ited by its primary focus on the magnitude of the deviation
as a sign of progression in intermittent exotropia. Although
the deviating angle tends to increase in many patients with
this disorder, other factors such as the patient’s control of
their deviation may be equally important in assessing pro-
gression. However, a notation of control was infrequently
recorded in patients’ medical records, and there is currently
no standard method for measuring control in this disorder.
In addition, the diagnosis and assessment of intermittent
exotropia was made by a variety of examiners within the
department, including resident and general ophthalmolo-
gists, which may have introduced some errors of measure-
ment. Finally, the ability to generalize the findings from this
population-based data to the United States as a whole is
limited by the lack of diversity in the population of Olmsted
County during the years of this study.
In this population-based cohort of children with intermit-
tent exotropia in which the deviation resolved in 4%, more
1158
than half were expected to have an increase of 10 or more
PD within 20 years, and 3 of 4 children were calculated to
undergo strabismus surgery within 20 years of diagnosis.
The results of this study, conducted in a defined population
for a 20-year period, although limited by uneven follow-up,
highlight the need for a prospective investigation of this
common disorder.
References
1. Mohney BG, Huffaker RK. Common forms of childhood
exotropia. Ophthalmology 2003;110:2093– 6.
2. Govindan M, Mohney BG, Diehl NN, Burke JP. Incidence and
types of childhood exotropia: a population-based study. Oph-
thalmology 2005;112:104 – 8.
3. Yu CB, Fan DS, Wong VW, et al. Changing patterns of
strabismus: a decade of experience in Hong Kong. Br J Oph-
thalmol 2002;86:854 – 6.
4. Hiles DA, Davies GT, Costenbader FD. Long-term observa-
tions on unoperated intermittent exotropia. Arch Ophthalmol
1968;80:436 – 42.
5. Kii T, Nakagawa T. Natural history of intermittent exotropia—
statistical study of preoperative strabismic angle in different
age groups [in Japanese]. Nippon Ganka Gakkai Zasshi 1992;
96:904 –9.
6. von Noorden GK, Campos EC. Binocular Vision and Ocular
Motility: Theory and Management of Strabismus. 6th ed. St.
Louis: Mosby; 2002:356 –76.
7. Rutstein RP, Corliss DA. The clinical course of intermittent
exotropia. Optom Vis Sci 2003;80:644 –9.
8. Chia A, Seenyen L, Long QB. A retrospective review of 287
consecutive children in Singapore presenting with intermittent
exotropia. J AAPOS 2005;9:257– 63.
9. Kurland LT, Molgaard CA. The patient record in epidemiol-
ogy. Sci Am 1981;245:54 – 63.
10. Pediatric Eye Disease Investigator Group. Spontaneous reso-
lution of early-onset esotropia: experience of the Congenital
Esotropia Observational Study. Am J Ophthalmol 2002;133:
109 –18.
11. Kaplan EL, Meier P. Nonparametric estimation from incom-
plete observations. J Am Stat Assoc 1958;53:457– 81.
12. Friendly DS. Surgical and nonsurgical management of inter-
mittent exotropia. Ophthalmol Clin North Am 1992;5:23–30.
13. Altizer LB. The nonsurgical treatment of exotropia. Am Or-
thopt J 1972;22:71– 6.
14. Cooper EL, Leyman IA. The management of intermittent
exotropia: a comparison of the results of surgical and nonsur-
gical treatment. Am Orthopt J 1977;27:61–7.
15. Newman J, Mazow ML. Intermittent exotropia: is surgery
necessary? Ophthalmic Surg 1981;12:199 –202.