June 2022: Results and Implications of the Convergence Insufficiency Treatment Trial–Attention & Reading Trial

Convergence Insufficiency Treatment Trial–Attention & Reading Trial Investigator Group. (2019a). Treatment of symptomatic convergence insufficiency in children enrolled in the CITT-ART randomized clinical trial. Optometry and Vision Science96(11), 825-835. https://doi.org/10.1097/OPX.0000000000001443

Convergence Insufficiency Treatment Trial–Attention & Reading Trial Investigator Group. (2019b). Effect of vergence/accommodative therapy on reading in children with convergence insufficiency: A randomized clinical trial. Optometry and Vision Science96(11), 836-849. https://doi.org/10.1097/OPX.0000000000001442

Summary by Jack M. Fletcher, Ph.D.


The Convergence Insufficiency Treatment Trial–Attention & Reading Trial (CITT-ART) is a multi-center randomized controlled trial (RCT) designed to evaluate whether optometric treatment of convergence insufficiency resulted in improved reading skills. Convergence insufficiency is a problem with binocular vision in which the eyes do not fuse when focused on stimuli in near vision. Children who are identified with convergence insufficiency complain of eye fatigue and discomfort, double vision, and headaches. When reading, children indicate difficulties concentrating, slow reading, and losing their place. Parents may enroll children in convergence insufficiency training in an effort to help them with reading. Some small-scale studies have reported improvement in reading after training. However, these studies had small samples, no control group, and other limitations.


The gold standard for determining the effectiveness of an intervention is an RCT. These two studies 2019a and 2019b report the results of an RCT designed to determine whether convergence training, delivered in an office context, led to improvement in reading skills. The trial involved 310 children 9-14 years of age identified with convergence insufficiency. Two-thirds of the children were randomly assigned to a treatment group in which the child received 16 weeks of convergence training. The other third was assigned to a placebo treatment that involved vision training with no binocular training. The placebo was designed to control for expectancy and attention effects associated with many interventions. Children in the placebo group received 16 weeks of vision therapy that did not address binocular vision or convergence issues. Pre- and postmeasures involved a primary outcome: reading comprehension. Secondary outcomes included measures of word reading, reading fluency, and other reading comprehension measures. There were also objective measures of binocular vision and convergence and symptom reports to assess whether the treatment resulted in improved convergence. Children with primary problems with foundational reading skills (e.g., dyslexia) were excluded from the study, which was wise since these types of vision problems should affect text reading, but not reading of words in isolation where saccadic eye movements are not required. The clinicians assessing outcomes were blinded to group assignment. Note that children were selected for convergence insufficiency and not for reading ability. Reading scores at baseline were slightly below national averages.


Study 2019a assessed whether the treatment led to improved binocular vision and a reduction in symptomatic complaints. This part of the study was important because if the treatment was not effective, then it would be difficult to conclude that any changes in reading were the result of convergence training. The results showed that on objective measures of binocular vision and fusion, the treatment group improved significantly more than the placebo group. Thus, the treatment was effective. However, there was no difference in reductions of symptomatic complaints, with both the treatment and comparison groups showing similar reductions in complaints. This attests to the impact of attention and expectancy on subjective complaints. There is clear evidence of improvement on objective clinical measures.

Study 2019b reports the results of the reading assessments. There were no significant differences in reading comprehension between the treatment and placebo groups. This null result extended to other reading comprehension measures and to measures of word reading and reading fluency. Thus, improvements in convergence and binocular vision are not associated with improved reading skills.


Convergence training reduces problems with binocular vision and fusion. It does not lead to improved reading skills. Practitioners should not suggest that a child undergo convergence training to improve reading skills. This conclusion is especially important for poor readers because resources should be devoted to reading intervention and not exercises that are not likely to generalize to the reading domain.


Behavioral optometry is controversial, with many proposing that vision exercises like convergence training will improve reading skills. This extends to activities like the use of colored overlays and changes in fonts (see Fletcher, Lyon, Fuchs, & Barnes, 2019, for a review of these hypotheses). When these interventions are presented as helping children with foundational reading problems, there should be no expectation of improvement in reading. Children with dyslexia have a primary problem reading and spelling words in isolation. Most hypotheses about the visual basis of dyslexia have to invoke a problem-reading text because these interventions at best affect the reading of connected text. Therefore, there should be no expectation of improvement in children with dyslexia because, as the authors of the trial note, dyslexia is a phonological/language problem. Colored overlays, altered fonts, and other visually-based exercises yield largely null results in appropriately controlled studies. More generally, the evidence for transfer from any intervention that does not directly train the skill of interest is limited, which helps explain why popularized notions like “brain training” do not lead to demonstrable improvements in academic skills (Moreau, 2022). More studies like these two RCTs are needed to determine whether these types of practices produce the hypothesized transfer outside of what is actually trained. 

More generally, there is little evidence that interventions that do not involve actual instruction in reading generalize to improved reading skills. This conclusion extends, for example, to training of cognitive skills. A recent review found little evidence that such training leads to improved reading skills (Moreau, 2022), adding to a growing body of evidence that such interventions may improve cognitive performance on tasks like those that are trained, but do not generalize to improved academic skills. For example, children with learning disabilities often have problems with memory. However, training memory skills with an intervention that does not involve reading, math, or writing does not lead to improved academic skills. Teachers, diagnosticians, and other practitioners should not refer struggling children for these types of interventions, especially if they detract from the resources available for academic interventions.


Fletcher, J. M., Lyon, G. R., Fuchs, L. S., & Barnes, M. A. (2019). Learning disabilities: From identification to intervention. Guilford Publications.

Moreau, D. (2022, April). How malleable are cognitive abilities? A critical perspective on popular brief interventions. American Psychologist77(3), 409-423. https://doi.org/10.1037/amp0000872