Comparison of the accommodative amplitude measured with and without the use of a specialised accommodative rule in children

Introduction

Accommodative amplitude measurement is an important recommended part of a routine eye examination.1 2 Accommodative amplitude plays a key role in the diagnosis and management of common refractive conditions such as presbyopia and latent hyperopia, as well as accommodative anomalies, especially accommodative insufficiency.3 4 Some ocular and systemic pathologies and also medications can affect the accommodative system, which can be detected by measuring the accommodative amplitude.5–7 Today, the importance of clinical evaluation of accommodative amplitude has been extended to modern ocular surgeries such as the implantation of accommodative intraocular lenses.8 In general, there are five methods for routine clinical assessment of accommodative amplitude, including push-up, push-down, push-down to recognition, minus lens and dynamic retinoscopy, with the majority being purely subjective.9 Open-view autorefractometers can provide a fully objective measurement of accommodative amplitude. However, their usage in optometric practice is limited, and their results are significantly influenced by pupil size.10–12 The push-up method is the most widely used and simplest measurement technique of accommodative amplitude. In this procedure, the patient is asked to look at a detailed target that is slowly approaching the eye and report whenever the first sustained blur occurs.13 14 To facilitate this testing, specialised tools called accommodative rules have been designed, including royal air force (RAF) rule, Berens Accommodation Rule and Krimsky Accommodation Rule. Although these devices have minor differences in design, their overall structure is such that they have a ruler-like body with a movable target on one side and the other side is placed in front of the patient’s eye.15 These instruments are commonly used in clinical practice as well as research.16 The advantages of these instruments include ease of use and the ability to effectively control the speed of the target movement as the precise gradation of intervals on the ruler visible to the examiner allows for effective regulation of the target’s approach towards the patient’s eye. On the other hand, the use of these devices is associated with the possibility of potential errors since the test conditions are more artificial compared with the real free space.17 18 However, it should be noted that the measurement of the accommodative amplitude with these accommodative rules is more popular in clinics, and most of the information related to the normal values of the accommodative amplitude that exists in the literature is based on this measurement protocol.16 19 Another way to measure accommodative amplitude is without an accommodative rule and instead by using a long millimetre ruler and a separate target.20 This method is more difficult to perform since the handling of the target and the ruler is done separately from each other. However, this method does not require special equipment and can be used in situations where there is no access to specialised accommodative rules.20 Also, measuring the accommodative amplitude with this method is closer to the person’s normal viewing conditions. According to the literature review, no study has examined the agreement between measuring the accommodative amplitude with and without the use of an accommodative rule. Such information is required to standardise the accommodative amplitude measurement method in both research and clinical practice. It is also necessary to answer the question of whether using accommodative rule for measuring accommodative amplitude is more reliable or not. The present study aimed to evaluate the agreement of accommodative amplitude measurements with and without an accommodative rule in children. This age group seems to be more important for such an assessment due to the more active accommodation.

Materials and methods

This report is part of the second phase of the Shahroud Eye Cohort Study,21 which was performed in 2018. All participants in the first phase of the study, who completed their examinations in 2015, were invited to participate in the second phase. After the individuals were present at the study site, first written consent was obtained from the parents and oral consent was obtained from the children. In the next stage, optometric examinations were done by the experienced examiners. The uncorrected distance visual acuity (VA) was measured using the Nidek CP-770 chart projector (Nidek, Gamagori, Japan) at 3 m. Objective refraction was performed by the ARK-510 A auto-refractometer (Nidek Co, Aichi, Japan). The autorefraction findings were refined with the retinoscope (Heine Beta 200 retinoscope, HEINE Optotechnic, Hersching, Germany). Subjective refraction was performed according to the maximum plus-best VA criteria to determine the optimal distance optical correction and the best-corrected distance VA was recorded.

In the next step, accommodative amplitude measurements were performed through the best optical correction by the two methods (with and without the accommodative rule). To evaluate the performance of the accommodative system more effectively, and avoid overestimating accommodative amplitude,12 the measurements were performed monocularly, first for the right eye and then for the left eye. To increase the measurement range, all accommodative amplitude measurements were performed through a −4 D lens, which was considered in the final estimation of the accommodative amplitude.

In measuring the accommodative amplitude using the Berens accommodative rule, the device was placed horizontally along the lower edge of spectacle plane. Then, the instrument’s target (English alphabetical letters equivalent to 20/30 VA) gradually approached the participant’s eye at a constant rate of 1 to 2 cm/s. The participant was instructed to keep the target as clear as possible and to report whenever the print became blurred and it was not possible to clear it with further effort (the first sustained blur). At this moment, the near point of accommodation (NPA) distance was recorded from the centimetre rating of the accommodative rule. To increase the reliability of the test, the NPA measurement was repeated three times. The NPA was then converted to accommodative amplitude in diopters (D) by using this formula: 100/NPA (cm).12 22 The average of these measurements was calculated for each participant as final NPA. A fixation stick containing English letters equivalent to 20/30 VA and a long millimetre ruler was used to measure accommodative amplitude without the specialised accommodative rule. This measurement was performed by the two trained examiners. The former was responsible for handling the target and the latter was responsible for measuring the distance. During the measurement (similar to measurement with the Berens rule), the target was slowly moved towards the participant at a rate of 1–2 cm/s. The participant was asked to try to keep the target as clear as possible and to report the occurrence of the first sustained blur. At this point, the NPA distance was measured from the spectacle plane. This measurement was repeated twice. The measurements were then converted to accommodative amplitude (D). Similar to the above, the average of these two measurements was recorded as the final NPA. The trained examiners first measured the three measurements with the Berens accommodation rule, and then the two measurements without the accommodation rule.

Finally, all participants underwent cyclo-refraction. The cycloplegic regimen included 1 drop of cyclopentolate 1% followed by 1 additional drop instilled 5 min apart, and cycloplegic refraction with the autorefractometer was done 30 min after the last drop. Refractive errors were defined according to the spherical equivalent of cycloplegic refraction. Myopia was defined as a SE ≤−0.50 D and hyperopia was considered as a SE ≥ +2.00 D. The exclusion criteria included a best corrected visual acqity (BCVA)worse than 20/30 in either eye, strabismus, amblyopia, a history of intraocular surgery, a history of ocular or systemic pathology affecting accommodation and ocular trauma.

Results

This study was performed on a subsample of 512 children who participated in the second phase of Shahroud Eye Cohort Study. After applying the exclusion criteria, 10 participants were excluded and finally, statistical analyses were done with data from 502 individuals and 1004 eyes. The mean age of participants was 11.7±1.3 years (range: 9–15 years) and 263 (52.4%) were male. Table 1 shows the mean±SD and mean differences of accommodative amplitude measurements with and without the use of the Berens accommodative rule by age, sex and refractive error. As seen in table 1, the mean measured accommodative amplitude without the use of the accommodative rule was 2.49±5.07 D higher than that measured with the Berens rule; this difference was statistically significant according to the paired t-test and in multilevel mixed effect analysis (table 2) (p<0.001). The mean difference of the measurements (with and without the Berens accommodative rule) was 2.30 D and 2.71 D in males and females, respectively (p=0.645). As seen in table 1, there was no clear trend in the mean difference of the measurements in different age groups. However, the highest and lowest mean difference was related to the 15-year and 10-year age groups, respectively. In terms of refractive errors, as shown in table 1, the highest and lowest mean difference was seen in emmetropes and hyperopes, respectively. Figure 1 illustrates the correlation of the accommodative amplitude measurements with and without the accommodative rule in the total sample. As shown in figure 1, the diagram has a funnel-like shape, and the agreement between the two methods decreases with increasing accommodative amplitude values. The overall ICC was 0.67 (95% CI 0.63 to 0.70). The highest and lowest ICC was related to the 10-year (0.73) and 13-year (0.54) age groups, respectively. Myopes had higher ICC compared with emmetropes and hyperopes, although the difference was not statistically significant. Figure 2 shows the Bland-Altman plot for the agreement of accommodative amplitude measurements by the two methods (with and without the Berens rule). It is clear in this figure that the variability of measurement errors is not uniform throughout the whole range of NPA measurements. Therefore, the Bland-Altman plot is misleading in this case.21 The differential and proportional biases were 2.844 and 0.765, respectively, which confirm the violation of assumptions for using the Bland-Altman plot. Figure 3 shows the amount of bias and emphasises that this bias increases at high values of NPA. In this regard, the amount of agreement between the two methods decreases dramatically at high values of NPA (online supplemental figure 1). The 95% limits of agreement were −12.5 to 7.5 in the total sample. The 95% limits of agreement were narrower in males compared with females (18.92 vs 20.87). The 95% limits of agreement were narrower in higher age groups and hyperopes (16.83 D) compared with emmetropes (18.37 D) and myopes (18.27 D).

Request permission

Request permission

Request permission

Table 2 shows that in a multilevel mixed effect model, NPA is 2.45 D (95% CI 2.16 to 2.73) higher when it is measured without an accommodative rule. Age, sex and refractive groups were not associated with NPA measurements. The ICCs for repeated measurements were 0.901 (95% CI 0.891 to 0.911) and 0.924 (95% CI 0.914 to 0.932) when measuring NPA with and without an accommodative rule, respectively. These results suggest that measurements without an accommodation rule have higher reliability.

Discussion

This study evaluated the agreement of accommodative amplitude measurements with and without the use of an accommodative rule in children. According to the results, the mean accommodative amplitude in the total sample measured with and without the Berens accommodative rule was 20.02 D and 22.46 D, respectively. Based on this, a difference of about 2.5 D was observed between the measurements of the accommodative amplitude with and without the use of the accommodative rule; this difference is clinically significant. Moreover, a moderate ICC and wide limits of agreement from −12.5 to 7.5 indicate a poor agreement between the two measurement methods. The agreement also was not constant and decreased in higher values of NPA. Various reasons can be proposed to explain this poor agreement between these two measurements. The first issue is the extent of control over the rate of movement of the target. Reaction time is a potential source of error in measuring the accommodative amplitude by the push-up method, which is affected by the speed of the target.18 25 The total reaction time itself is a sum of four reaction times, which are the time it takes for the patient to perceive the first sustained blur, the time interval from the moment of perception to the report of the blur, the time it takes for the examiner to receive and interpret the patient’s response and eventually stop the target’s movement.16 The reaction time error increases with increasing target velocity nonlinearly.16 18 Accordingly, it is recommended to move the target towards the patient’s eye slowly and steadily while performing the push-up test. The difference between the two methods can also be due to the effect of the proximal cues on the accommodative amplitude measurements by the push-up method. The accommodative response is regulated by three signals: retinal, convergence and psychological.26 One psychological factor that can affect accommodation is the patient’s awareness of the nearness of the test target.27 28 In accommodative amplitude measurement using an accommodative rule, the contact of the instrument with the patient’s face provides proprioceptive feedback.29 This feedback may increase the awareness of proximity, in which case the person feels that the target is closer than its actual distance.29 This provides a stimulus to cease accommodative effort early that ultimately leads to a more remote NPA (less accommodative amplitude). The different effects of the depth of focus on the accommodative amplitude measurements due to the variations in target parameters (luminance, sharpness, shape and contrast)30 can also be considered as a potential factor for the difference in the measured accommodative amplitude with and without the use of an accommodative rule.

In the present study, subgroup analysis was done to assess the agreement of accommodative amplitude measurements with and without the use of a specialised accommodative rule in different groups of sex, age and refractive errors. According to the results, the accommodative amplitude values measured without the use of the accommodative rule were higher than the measurements by the rule in both sexes. However, the mean difference in measurements was less in males versus females (2.30 vs 2.71 D). Furthermore, the limits of agreement were approximately 1.95 D narrower in males. Although this sex difference indicates a better agreement of measurements in males compared with females, this difference was not statistically significant and does not appear to have a clinical significance. The results of this study showed that in all age groups, the mean measured accommodative amplitude without the accommodative rule was significantly higher than the mean accommodative amplitude measured with the Berens accommodate rule. The mean difference of measurements did not show a clear pattern with age; however, the agreement was narrower in higher age groups. This finding suggests that the agreement of the measurements between the two methods improved with increasing age. One possible explanation for this finding could be an increase in patient cooperation due to a better understanding of the test conditions and instructions. Also, given that at a younger age, the accommodative amplitude is higher and the NPA is closer,31 overestimation or underestimation of distance will have a more significant effect on the final measured accommodative amplitude. For example, at a distance of 10 cm, each centimetre of change is equal to 1 D, but at a distance of 5 cm, it will be equal to 5 D. This is well illustrated in figures 1 and 3 and supplemental figure 1. As shown in figure 1, the diagram has a funnel-like shape, and the agreement between the two methods decreases with increasing accommodative amplitude values.

According to the results, the mean accommodative amplitude without the accommodative rule was significantly higher than the accommodative rule’s mean accommodative amplitude in all three refractive groups. The mean difference of measurements in hyperopes was less than in other refractive groups. Hyperopes also had a higher ICC and narrower limits of agreement than emmetropes. These findings suggest a better agreement between the measurements of the accommodative amplitude in hyperopes; however, this difference between refractive groups was not significant in the multilevel mixed effect model, and by accounting for all sources of correlation between repeated measures, individuals and their eyes. Various factors can be mentioned in explaining these differences, including differences in the depth of focus32 and the blur perception,33 34 as well as accommodative variations in different refractive groups.

One of the strengths of this study was the examination of the agreement of two common methods of measuring NPA with a large sample size. Proper statistical analysis and the use of new methods for detecting bias and agreement were other strengths of this study. In fact, in this study, using the traditional Bland and Altman method for reporting agreement is misleading. Instead, the bias was reported by using new method, introduced by Taffé et al.23 However, the number of measurements in the two methods was not the same. This discrepancy may have reduced the measurement accuracy in the method that did not use an accommodative ruler. Additionally, NPA measurement was conducted initially with an accommodative ruler, followed by measurements without a ruler. The non-random sequence of NPA measurement methods could be seen as a limitation of this study, as it may have influenced the responses of the students, potentially leading to more accurate answers in the fourth and fifth measurements. It was impossible to mask the examiners from the previous measurements, which can be seen as another limitation of this study. However, it is important to note that they were trained and continuously supervised by the principal investigator.

In conclusion, the findings of this study showed poor and non-constant agreement between the measurements of the accommodative amplitude with and without the use of an accommodative rule. The agreement of the two methods was better in lower amounts of accommodative amplitude and older ages. Given the higher reliability of measurements, it can be concluded that measuring NPA without an accommodative rule, which is also recommended in the binocular vision textbooks,20 is more appropriate. These issues should be considered in clinical measurements of accommodative amplitude in children.