A pinboard by
Elise Harb

Clinician Scientist, UC Berkeley School of Optometry


Does your behaviors cause you to need glasses?

Subjective and Objective Measures of Human Behavior Are Not Equivalent EN Harb, M Guillen, JE Choi, M Chan, Z Sajjadi, CF Wildsoet

Purpose: Reports suggest that near work may be a risk factor for and outdoor activity protective against the development of myopia. In this study, the reliability of traditional questionnaires to measure habitual human behavior was assessed by comparison with objectively measured activities.

Methods: UC Berkeley students aged 18-24 (n=10 non-myopes, SER range: +1.63 to -0.63 D; n=19 myopes, -1.38 to -6.25 D) participated. Subjects wore continuously recording light sensor devices (Actiwatch) on their non-dominant arm over clothes for 2 weeks. Data were averaged over 1-min epochs, with white light intensity ≥1000 lux used as a proxy for time outdoors. As a measure of near work, subjects were prompted to estimate hours of near work every evening by the device. A subjective questionnaire about habitual activities was also completed by each subject at the start of their wearing period and once during the 2-week period, with responses from the two questionnaires averaged for use in analysis. Questionnaire reliability, as compared to objective Actiwatch data, was assessed for academic (AP) and/or non-academic (NAP) periods (n=14 measured during both periods).

Results: Overall, subjects misreported time spent indoors and outdoors, with individuals varying widely in the accuracy of their estimates, especially in the NAP (Questionnaires – Actiwatch: +/- 300 min outdoors vs. +/- 400 min indoors). Similarly, subjects misreported their near work activities, especially in the AP (+/- 200 min AP vs. +/- 100 min NAP). The nature of reporting errors also appeared to vary between the two groups. For example, for NAP, myopes overestimated by ~20% and non-myopes underestimated by ~30%, time spent outdoors (p=0.03), and myopes tended to underestimate indoor activity more than non-myopes (~10% diff, p=0.07). For AP, myopes generally underestimated and non-myopes overestimated near work activities.

Conclusions: These results demonstrate the unreliability of questionnaire data in capturing the habitual behaviors of young adults. Given the further possibility that reporting errors may differ between myopes and non-myopes, the use of objective measures of habitual activity in studies aimed at advancing our understanding of the role of behavior in myopia development and/or progression would seem imperative.