Projected generational increase in myopic retinopathy in the United States

Abstract

Purpose: Vitale et al (2009) used NHANES data to estimate a 66% increase in myopia prevalence in the United States (US) between 1971-72 to 1999-2004. The aim of this work is to predict the likely impact of this generational increase in myopia on the future prevalence of myopic retinopathy (MR).

Methods: In the absence of MR prevalence data at different refractive errors (RE) in US-based studies, 3 international population-based studies were used: the Blue Mountains Eye Study (BMES), the Beijing Eye Study (BES) and the Hisayama Study (HS). MR prevalence at a given RE varied greatly between these studies; however simple exponential regression of the MR odds ratios versus RE (y = αe^-βx, where y is the odds ratio and x is RE) yielded a relatively consistent slope, -β, in semi-log plots across studies. Relative inter-generational MR prevalence was predicted by multiplying the average MR odds ratio at each RE (obtained from the average β) with the corresponding myopia prevalence data for both of the 1971-72 and 1999-2004 NHANES data sets and then summing across each set. Multiple assumptions were necessary in producing the estimate, including the following: myopia prevalence at age 12-54 years is assumed to be predictive of MR prevalence some 30 years later in those above 40 years of age; the BES study did not provide a denominator in odds-ratio estimates, so a dummy estimate was used; open-ended refractive error bins were assigned representative estimates.

Results: As noted above, the lines of best fit relating refractive error and the log of the odds ratios for each MR prevalence study were excellent fits and of surprisingly similar slopes (BMES; R²=0.99, β=0.61: BES; R²=1.00, β=0.67:HS; R²= 0.97, β=0.61), supporting use of non-US MR prevalence odds for this analysis. There was no threshold apparent between low (physiological) myopia and high (pathological) myopia. The average increase in the odds of MR for each diopter increase in myopia is estimated to be 88%. The predicted increase in the overall prevalence of MR for the US is 3.0 times.

Conclusions: While the estimate remains subject to multiple assumptions, there is a disturbing projected threefold increase in MR prevalence expected by about 2030 compared to that at the turn of the century.

Keywords: 605 myopia • 463 clinical (human) or epidemiologic studies: prevalence/incidence