In Reply We agree with Dr Zhang and colleagues that age standardization is important when estimating national prevalence trends. In our methods for estimating prevalence, we included age as one of the essential covariates in the multivariable regression models.1
Although joinpoint regression to detect changes in trends of prevalence over time is used in some studies, this methodology is more applicable to investigation of nonlinear quadratic trends and when the quadratic or higher-order time variable in the regression model is statistically significant. Since one of the primary objectives of the study was to determine whether there was an overall statistically significant linear trend over time from 2001 to 2016, we did not consider the quadratic trend factor to warrant reporting. However, we have now conducted these analyses testing for quadratic trends and the results were not significant (P = .38). While many potential methods can be used to conduct trend analyses, each with its own limitations, our decision to use logistic regression with orthogonal polynomials was based on its applicability to cross-sectional survey–based study cohorts and existing studies that demonstrate these methods to be reliable and accurate for analyzing linear trends in complex survey data.2,3
Dang K, Wong RJ. Trends in Alcoholic Fatty Liver Disease—Reply. JAMA. 2019;322(10):980–981. doi:10.1001/jama.2019.10351
Browse and subscribe to JAMA Network podcasts!
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: