Factors associated with base-in prism treatment outcomes for convergence insufficiency in symptomatic presbyopes


Dr Yi Pang, Illinois College of Optometry, 3241 S Michigan Avenue, Chicago, Illinois 60616, USA, E-mail: ypang@ico.edu


Purpose:  The aim was to determine whether baseline measurements of binocular vision are associated with symptoms of convergence insufficiency (CI) both before and after CI treatment with base-in prism in symptomatic presbyopes. We further determined the factors that predicted CI treatment outcomes.

Methods:  A total of 29 symptomatic CI subjects aged 45 to 68 years were examined and the following binocular vision measurements were obtained at the baseline: near heterophoria by cover test, associated phoria at near (AP), near point of convergence (NPC) and positive fusional vergence at near (PFV). Each subject wore each of two pairs of progressive addition spectacles for a three-week period, one with prism and one with the same prescription but no prism. The wear order was randomised across subjects. Subjects completed the CI symptom survey (CISS) before and at the end of the three weeks, thus baseline, placebo and treatment CISS scores were tabulated. The reduction in symptoms (CISS score improvement) was calculated as the difference between the placebo and treatment CISS scores.

Results:  Neither baseline CISS score nor treatment CISS score had statistically significant correlations with any of the binocular vision measurements. CISS score improvement was statistically significantly associated with near heterophoria, AP and NPC break but not with PFV. CISS score improvement was significantly predicted by both near heterophoria and NPC break but not by AP and PFV. With other variables held constant, the CISS score improved by 1.57 for each additional prism dioptre of exophoria at near and by 0.81 for each additional centimetre that the NPC break was receded.

Conclusion:  There was no significant association between the CISS score with any of the baseline binocular vision measurements before or after prism treatment. Our findings suggest that a greater near heterophoria and/or more receded NPC before treatment are associated with more reduction in CI symptoms by prism treatment.

Convergence insufficiency (CI) is one of the most common binocular vision disorders and has been reported to have a prevalence of 2.5 to 25 per cent.1–4 It can cause headache, blurred vision, visual fatigue, double vision and other symptoms.5,6 Several symptom surveys have been developed to evaluate CI.5,7–10 Among all the instruments, the CI symptom survey (CISS) is most commonly used11–13 and has been validated both in children and young adults.5,8 A CISS score of 16 or more has been confirmed to be a good discriminator of CI from normal binocular vision in children aged nine to 18 years,5 while Rouse and colleagues8 recommended a cut-off value of 21 for CI in young adults aged 19 to 30 years.

The success or failure of CI treatment is determined by changes in clinical signs as well as changes in patient symptoms.5 Several studies show that symptoms of CI can be reduced by vision therapy and other CI treatments;11,12,14 however, the reduction in symptoms varied with treatment. The Convergence Insufficiency Treatment Trial Study Group14 studied the effect of office-based vergence/accommodative therapy and reported that 73 per cent of the children studied had an improvement of more than 10 in their symptom scores or CISS less than16 following vision therapy, while the rest of the subjects had less reduction in their symptom score. Clinically, it would be helpful to be able to predict the level of reduction in symptoms before CI treatment.

Teitelbaum, Pang and Krall13 confirmed that base-in (BI) prism was effective in alleviating symptoms of CI in presbyopes. The purpose of the present study was to determine whether certain binocular vision measurements tested at baseline were associated with symptoms of CI both before and after base-in prism treatment in symptomatic presbyopes. In addition, we investigated the factors that predicted the outcomes of CI treatment.


The study protocol and informed consent forms were approved by the Illinois College of Optometry Institutional Review Board and followed the tenets of the Declaration of Helsinki. Informed consent was obtained from all subjects in the study. The design of the study has been published in a separate article13 and is only partially described below.

Twenty-nine subjects from a private optometric practice in South Dakota were recruited for the study. Potential subjects aged 45 years or older were asked to complete the Convergence Insufficiency Symptom Survey-V15 (CISS-V15). In previous studies with subjects aged nine to 18 years5 and 18 to 30 years,8 the survey was administered by reading to the subjects. Because our subjects were adults, the survey was self-administered. A comprehensive eye examination including refraction was performed for the patients whose CISS-V15 score was higher than 16. Because in clinical practice treatment of CI is commonly provided only to patients with symptoms, only the symptomatic subjects were enrolled in the present study to determine the factors associated with CI treatment outcomes. To qualify for the study, subjects had to have an exophoria at near of at least 4.0Δ greater than at distance. The following diagnostic data were collected by one of the authors (JK):

distance heterophoria and near heterophoria by cover test

associated phoria (AP) at near measured with the Bernell binocular refraction slide

the near point of convergence (NPC) using an accommodative target

positive fusional vergence (PFV) at near measured with Risley prisms.

Eligibility criteria are listed as follows:

visual acuity of 6/7.5 or better in each eye at distance and near

currently wearing progressive addition lenses

a minimum of +1.50 D addition in the subject's habitual prescription

a minimum of two hours spent on reading or close work on a daily basis

AP at near of at least one prism dioptre base-in

no AP with the potential base-in prism at distance

exophoria at near at least 4.0Δ greater than at distance

CI symptom score of 16 or more.

Subjects with constant strabismus at distance or at near, vertical heterophoria greater than 1.0Δ or a history of treatment with prism were excluded from the study. The three levels of CI classification created by the CITT Study Group were used:3,15

CI-1 refers to a greater exophoria at near than at distance by at least 4.0Δ

CI-1 with one additional clinical sign (a total of two signs) is classified as CI-2

CI-1 with two additional clinical signs (a total of three signs) is categorised as CI-3.

Additional clinical signs include receded NPC (6.0 cm or more to break) and insufficient PFV by either failure to reach Sheard's criterion or failure to demonstrate minimum normative PFV at near (up to 15 to PFV break). There were no subjects with CI-1, six subjects with CI-2 and 23 with CI-3 in the present study.

Subjects were assigned two pairs of progressive addition lenses in a randomised sequence. The prism was prescribed based on the neutralisation of the AP at near in one pair of glasses, while the other pair served as a placebo. The prism was incorporated in a novel progressive addition lens where the prism was confined only to the near portion and equally split between two eyes (Kodak lenses with anti-fatigue progressive technology, http://www.signetarmorlite.com). The maximum amount of prism that could be incorporated in the near portion only ranged from 0.375Δ to 0.75Δ per lens in the near portion and varied according to the base curve and power of the lens. Due to these limitations, additional prism (mean and SD: 0.57 ± 0.63Δ per lens), when required, was incorporated throughout the entire lens in 20 subjects to ensure that the effect of the prism through the distance portion of the glasses was minimised and potentially did not contribute to subjective symptoms and the CISS score. The AP at distance had to be zero for subjects to be enrolled in the study. A double-blind design was used so that neither the examiner nor the subject was aware of the assigned spectacles. Subjects wore each pair of glasses for three weeks and completed the CISS-V15 at the end of each wearing period. Therefore, a placebo and treatment CI symptom score were tabulated in addition to the baseline score. Because a placebo effect was detected (the CISS score in the placebo condition was significantly different from the baseline),13 the reduction in symptoms with prism (CISS score improvement) was calculated as the difference between the placebo and treatment CISS scores.

The Kolmogorov-Smirnov test was used to check the normality of distributions of the measurements of binocular vision and baseline, placebo and treatment CISS scores. The AP was the only parameter that was not normally distributed. Reciprocal transformation was performed, which resulted in a normal distribution. None of the parameters was an outlier, which was confirmed by the analysis of Cook's distance, Mahalanobis distance and standardised residual. Statistical sample size calculation was performed to determine the number of subjects required in the study to detect differences in CI symptom scores. A sample size of 21 would be required to give 80 per cent power at the 0.05 level and 28 subjects were needed to give 90 per cent power. Therefore, a sample size of 29 subjects was determined to be adequate for the present study. The correlation of binocular vision measurements with CISS score was evaluated using the Pearson correlation. Multiple regressions with backward analysis were used to identify which binocular vision measurements predicted CISS score improvement.16


The baseline characteristics of the subjects and clinical measurements were described in our previous paper.13 The average age (and SD) of the subjects was 54.14 ± 5.90 years. The mean exophoria at distance was 2.24 ± 1.75Δ and at near was 11.17 ± 3.49Δ. The mean AP at near was 2.29 ± 1.31Δ. The mean addition power was 2.04 ± 0.30 D. The mean baseline, placebo and treatment CISS score was 30.21 ± 9.30, 23.62 ± 10.76 and 13.38 ± 9.44, respectively.

Neither baseline nor treatment CISS score had statistically significant correlations with any of the binocular vision measurements measured at enrolment (Table 1). The CISS improvement score was statistically significantly associated with near heterophoria, AP and NPC break but not with the PFV break (Table 1 and Figure 1). Multiple regression analysis showed that both the near heterophoria and NPC break significantly predicted the CISS score improvement but not the PFV break (Table 2). The AP did not add extra predictability in the presence of near heterophoria and NPC (p = 0.09, Table 2). With other variables held constant, the CISS score improved by 1.57 (coefficient) for each additional prism dioptre of heterophoria at near and by 0.81 (coefficient) for each additional centimetre that the NPC break receded (Table 2). Using the criterion suggested by the CITT Study Group that a reduction of 10 in the CISS score signifies clinical relevance,14 we recommend that six prism dioptres of exophoria (=10/1.57) and/or 12 cm NPC break (=10/0.81) is the threshold for predicting successful prism treatment outcomes in symptomatic presbyopes with CI.

Table 1. Correlation coefficients (r) between binocular vision measurements at near and CISS score
CISS scoreNear HeterophoriaAPNPC breakPFV
  • Statistical significance at the p < 0.05 level.

  • ** 

    Statistical significance at the p < 0.01 level.

  • AP: associated phoria, CISS: convergence insufficiency symptom survey, NPC: near point of convergence, PFV: positive fusional vergence

Baseline CISS score0.100.160.13−0.24
Treatment CISS score0.05−0.04−0.12−0.03
CISS score improvement−0.45**−0.39*0.42*−0.09
Figure 1.

The correlation between convergence insufficiency symptom survey (CISS) score improvements and binocular vision measurements. A CISS score improvement greater than zero indicates an improvement of symptoms of convergence insufficiency with base-in prism. A CISS score improvement less than zero indicates symptoms of convergence insufficiency deteriorated with base-in prism. NPC: near point of convergence, PFV: positive fusional vergence

Table 2. Multiple regression analysis for factors predicting convergence insufficiency symptom survey score improvement with base-in prism
Binocular vision measurement at nearCoefficientR2p-value
  • Statistical significance at the p < 0.05 level.

  • AP: associated phoria, NPC: near point of convergence, PFV: positive fusional vergence

Near Heterophoria1.570.210.01*
NPC break0.810.180.02*
PFV break0.0010.011.00


In the present study we found that the CISS score improvement by prism treatment was statistically significantly correlated with near heterophoria, AP and NPC. Interestingly, while baseline and treatment CISS scores did not correlate with any of the binocular vision measurements, both the near heterophoria and the NPC significantly predicted the prism treatment outcomes indicated by CISS score improvement.

Although the AP is not the most common method used clinically to prescribe horizontal prism, it has been shown to be a better indicator of near symptoms in presbyopes compared with dissociated heterophoria17,18 Yekta, Pickwell and Jenkins18 tested 118 pre-presbyopes and 69 presbyopes and found that the AP was related more to binocular vision problems at near compared with heterophoria and the forced vergence disparity curve. In addition, Jenkins, Pickwell and Yekta17 determined that the AP was more reliable for predicting the near symptoms than heterophoria both in pre-presbyopes and presbyopes. Thus, the amount of prism prescribed in the present study was based on the neutralisation of AP at near. Previous studies have used two to six weeks to evaluate glasses performance.19–22 In the present study, a period of three weeks was chosen for subjects to wear each pair of glasses, so that our subjects had enough time to experience the different pairs of glasses and the wearing period was not too long to affect wearing compliance.

Studies examining the relationship between binocular vision measurement and symptoms have yielded conflicting results.9,10,15,18,23,24 Delgadillo and Griffin23 studied third and fourth year non-presbyopic optometric students. They did not find any difference in vergence facility between students with or without symptoms. Borsting and colleagues15 studied children in grades three to eight and found CI symptom scores had a significant correlation with NPC but not with PFV and near heterophoria. Sheedy and Saladin9 studied 38 pre-presbyopic subjects with exophoria and found that heterophoria was an important discriminator between symptomatic and asymptomatic subjects. They determined that PFV was also a significant discriminator but not as strong as heterophoria. Yekta, Pickwell and Jenkins18 studied 187 subjects aged 10 to 65 years and reported that symptoms were significantly correlated with AP and fixation disparity but not with heterophoria.18 NPC and PFV were not evaluated in their study.

Unlike those in previous studies, the subjects in the present study were all patients with symptomatic CI (both CI-2 and CI-3). The severity of symptoms, either before or after CI treatment, was not found to be associated with any of the binocular vision measurements. This finding could be due to a couple of factors. First, most of the CISS questions depend on subjective judgements from patients. For example, the question ‘Do your eyes feel uncomfortable when reading or doing close work?’ is based on the subjective impression of the subject and similar levels of discomfort might be scored differently. Second, although our subjects were requested to perform near work for at least two hours per day, the duration of close work might have varied among subjects. Subjects who read more might complain more than subjects who read less. By using the CISS score improvement, the variability in subject response to the severity of symptoms was partially mitigated as relative changes in symptoms by subjects might be more accurate than absolute scores where no detailed descriptions were attached to the survey scoring. This might explain why only CISS score improvement showed a significant correlation with binocular vision measurements.

CI is characterised by one or more of the following three clinical signs, namely, exophoria at near that is greater than at distance, decreased PFV and a receded NPC.1,4 The findings in the present study suggest that a greater near heterophoria and/or more receded NPC are associated with more reduction in symptoms of CI by prism treatment while PFV had no significant correlation. Due to the limited range of AP values in the current subjects, it would be unlikely to find a strong association between AP and CI symptom score. Previous studies suggest that PFV has low repeatability.25–28 Rouse, Borsting and Deland28 tested the repeatability of PFV with Risley prisms one week later in 20 children aged 10 to 11 years and found that PFV was less repeatable compared with near heterophoria and NPC. They concluded that clinicians could measure differences in PFV as large as 12Δ on different visits. The low repeatability of PFV might explain why no significant correlation was identified between CISS score improvement and PFV in the present study. In contrast, both near heterophoria and NPC have been shown to have high reliability and repeatability.28

There are a few limitations in the present study. First, while a CISS cut-off value of 16 has been determined in children5 and suggested to be 21 in young adults,8 no recommendations have been made for the presbyopic population.8 Furthermore, Sheedy and Saladin24 have shown that presbyopes are less symptomatic than non-presbyopes. Therefore, a cut-off value of 16, which is lower than the value in young adults, was chosen to reflect that presbyopes are less symptomatic. A study is warranted to determine the CISS cut-off value in presbyopes. Second, because the subjects were symptomatic presbyopes, the results of the present study cannot be generalised to children and the young adult population and they cannot be generalised to an asymptomatic CI population. Third, base-in prism was used to treat CI subjects in the present study and the results cannot be generalised to other CI treatments.


Our findings suggest that neither baseline CI symptoms nor post-treatment symptoms are associated with any of the binocular vision measurements (near heterophoria, NPC, AP and PFV) in symptomatic presbyopes with CI (both CI-2 and CI-3) but a greater near heterophoria and/or more receded NPC before treatment are associated with larger reductions in symptoms of CI after prism treatment. This might help the clinician advise symptomatic CI patients on the expected outcomes of treatment.