An evaluation of the lag of accommodation using photorefraction.

Objective: The lag of accommodation which occurs in most human subjects during reading has been proposed to explain the association between reading and myopia. However, the measured lags are variable among different published studies and current knowledge on its magnitude rests largely on measurements with the Canon R-1 autorefractor. Therefore, we have measured it with another technique, eccentric infrared photorefraction (the PowerRefractor), and studied how it can be modified.

Methods: Particular care was taken to ensure correct calibration of the instrument. Ten young adult subjects were refracted both in the fixation axis of the right eye and from the midline between both eyes, while they read text both monocularly and binocularly at 1.5, 2, 3, 4 and 5 D distance ("group 1"). A second group of 10 subjects ("group 2"), measured from the midline between both eyes, was studied to analyze the effects of binocular vs monocular vision, addition of +1 or +2 D lenses, and of letter size. Spherical equivalents (SE) were analyzed in all cases.

Results: The lag of accommodation was variable among subjects (standard deviations among groups and viewing distances ranging from 0.18 to 1.07 D) but was significant when the measurements were done in the fixation axis (0.35 D at 3 D target distance to 0.60 D at 5 D with binocular vision; p<0.01 or better all cases). Refracting from the midline between both eyes tended to underestimate the lag of accommodation although this was significant only at 5 D (ANOVA: p<0.0001, post hoc t-test: p<0.05). There was a small improvement in accommodation precision with binocular compared to monocular viewing but significance was reached only for the 5 D reading target (group 1--lags for a 3/4/5 D target: 0.35 vs 0.41 D/0.48 vs 0.47 D/0.60 vs 0.66 D, ANOVA: p<0.0001, post hoc t-test: p<0.05; group 2--0.29 vs 0.12 D, 0.33 vs 0.16 D, 0.23 vs -0.31 D, ANOVA: p<0.0001, post hoc t-test: p<0.05). Adjusting the letter height for constant angular subtense (0.2 deg) induced scarcely more accommodation than keeping letter size constantly at 3.5 mm (ANOVA: p<0.0001, post hoc t-test: n.s.). Positive trial lenses reduced the lag of accommodation under monocular viewing conditions and even reversed it with binocular vision.

Conclusions: After consideration of possible sources of measurement error, the lag of accommodation measured with photorefraction at 3 D (0.41 D SE monocular and 0.35 D SE binocular) was in the range of published values from the Canon R-1 autorefractor. With the measured lag, simulations of the retinal images for a diffraction limited eye suggest surprisingly poor letter contrast on the retina.