Previous studies have evaluated pupillary size in both ambient photopic (4, 5, 11, 14, 15) and scotopic conditions (12, 14, 15). There has also been assessment of pupillary function in ambient light conditions alone (4, 5, 11). Table 3 shows the results of pupillary size and function from the aforementioned studies. This study is the first to propose normative values for parameters of scotopic pupillary function in a pediatric population in typical clinical conditions.
It is well established that, throughout childhood, the eye and visual pathway have significant growth and development, which may expectantly lead to pupillary changes over time. Although under ambient light conditions, Brown et al. measured data similar to our results that suggested a weak, yet statistically significant, correlation between pupil size and age (resting diameter, r=0.29; post-stimulus aperture, r=0.19). Other studies in light (11) and dark conditions (12) suggest a peak in pupil size around 11 years old, which was shown in light to continue to decrease throughout adulthood.(16) However, we did not see a similar peak and subsequent decrease in our scotopic pediatric data. Our contrast in findings compared to Kohnen et al. might be attributable to differences in population sampling; ours included a larger number of subjects, especially in ages above 11 years old.
In collecting pupillometry data, the device used reports data in two colors: green (reliable) and red (unreliable). It will only display as green if it generates reliable data in at least six out of the eight variables measured. This explains the discrepancies in n values of ADV and T75 from the remainder of the data. Adhikari et al. showed that pupil diameter measured by an infrared camera is affected by fixation eccentricity, introducing at least a 0.07-millimeter error once eye movements deviate more than 5º from the central fixation axis. (17) While fixation eccentricity was not measured in this study, the authors felt the levels of reported reliability from the pupillometer served as an appropriate surrogate when considering the nature of examining children in a clinical setting. Multiple studies (4, 11, 12) have mentioned the inherent difficulties in measuring children younger than 5 years old. We found this to be true for practical purposes, as well; 28/42 of the patients unable to cooperate with the exam were less than 5 years old. Among these children, the most encountered difficulties were fear of the pupillometer, inconsolable crying, and constant head movement. The remaining 14/42 patients were between 5 and 10 years old. These children cooperated well with the exam, but they were unable to avoid blinking during the pupillary constriction period of measurement — thus preventing adequate data collection. With more experience, younger children were able to be distracted for sufficient periods of time to collect data, and older children tolerated assistance from the examiner’s hand to prevent eye blinking. Thus, proper training and repeated use, just as with other instruments used in eye examination, can yield more quantitative measurements that were otherwise subjectively based. Although we presented a smaller sample size for younger individuals, we hope to provide some amount of normative comparison in that age group. Finally, for future study, a larger sample size can strengthen the significance of a normative database.
Previous studies have been equivocal on the association between gender and pupillary reaction, showing weak differences (14) or no difference at all (11). Our data show statistical differences only between the constriction percentage and maximum constriction velocities of young males and females (p=0.0179 and p=0.0105, respectively). In contrast to Fan et al., our findings might be attributable to differences in population sampling: ours having a higher n and lower ages than their population.(14) Our data suggest that a potentially abnormal patient should be compared to one of the same gender from our data.
Via self-reporting, demographic data was obtained for all subjects on ethnicity but for only 49/101 subjects with regards to race. There were no statistical differences for any measured variable by ethnicity or by race. However, Brown et al. measured a population that was 64% white and 24% African American, finding statistical differences in pupil size and function (MCV and ADV) between the eyes of White and African American subjects when measure in ambient light. Due to differences in demographics between the two studies, direct comparisons cannot be made.
Anisocoria has been clinically defined as asymmetry of 1 mm or greater (4, 5). Interstimulus interval between eyes has been shown to affect measurement of pupillary light response amplitudes if the pupil has not returned to its resting diameter, and the necessary interval is shorter in scotopic conditions.(18) Order of eye measurement, thus, could affect anisocoria data if the interstimulus interval is not appropriate. During this study’s data collection, the interval between eye measurements was roughly one minute while the previous eye data was transcribed. There was no significant trend found in the data to favor the resting diameter of one eye to be greater than the other. As shown in Table 3, multiple studies including the present one show the majority of patients to have pupillary asymmetry of less than 0.5 mm on quantitative pupillometry. Therefore, the authors recommend future investigation into detection of pathology by using thresholds of 0.5 mm and 1 mm for anisocoria to determine the potential utility of quantitative pupillometry over standard subjective measurement as a screening tool as well as determining exact sensitivity, specificity, and receiver-operator curves for each method using these thresholds.
The clinical nature of this study provides both strengths and limitations to the data and analysis presented above. As mentioned earlier, fixation eccentricity and exact interstimulus interval provide a more standardized approach to data collection. Furthermore, data were collected throughout the clinic day, thus preventing consistent time measurements to limit effects from circadian rhythm on pupillary function.(19) Finally, given the clinical setting of these measurements, children are likely to have many psychosensory influences on pupil size and function such as anxiety, fright, and pain, all of which would exhibit mydriatic effects on the data. However, the authors wish to highlight that when clinically evaluating a pediatric population, precise time of day and measurements that require cooperation as well as the psychosensory effect of visiting any clinical setting whether it be in a hospital-based clinic or an outpatient clinic are rarely standard. Therefore, the data presented in this study provide a robust database with which comparisons can be made with a pediatric population in a clinical setting.