To consider the change in prevalence longitudinally, the 95% confidence intervals of the prevalence were calculated by the Wald method and were considered significantly different if the confidence intervals for each percentage did not overlap. Cross-sectional association with age was considered with univariate logistic regression. The prevalence of the following independent variables were first considered in a logistic univariate analysis: age, sex, better and worse eye corrected VA, presence of visual impairment, presence of low vision, comorbidity score, ocular disease score, smoking, and use of antidepressants. All univariate and multivariate analyses were undertaken on the data collected in the same time period (i.e., not between the current and past datasets). The dependent variables were the presence of either BV or eye movement AT, disorder in the current data or 10 years prior. For each of these dependent variables, a multivariate analysis was undertaken using forward stepwise logistic regression, with alpha to enter equaling 0.05 and alpha to remove equaling 0.15. The independent variables that were included were those that reached significance (P < 0.05) or were moderately close to significance (P < 0.15) in the univariate analysis. Univariate analysis was also used to determine any association with age for each of the individual past and current ATs and disorders (e.g., presence of strabismus or large distance exophoria). Data were analyzed in a spreadsheet application (Microsoft Excel; Microsoft Corp., Redmond, WA) for the descriptive statistics and a statistics software package (Systat; Systat Software, Inc., Chicago, IL) for the regressions. For significance, a value of P < 0.05 was used.