05-31-2018 12:29 PM
I am now trying to run the trend analysis from the NHANES survey data. In general data, the trend analysis are simple for continuous variable, which can be run like
proc freq data=data; table CycleYear*age/JT; run;
and categorical variable, which can be run like
proc freq data=data; table Diabetes*CycleYear/trend; run;
I am just wondering for survey data like NHANES data, how can I run this similar trend analysis for both continuous and categorical variable? For example of categorical variable "diabetes Yes vs. No", I want to know the Diabetes rate change or not from different survey cycle year like from 2001-2002, 2003-2004, 2005-2006, .... 2015-2016. there was total 8 cycle of survey data. if there was a trend of increase/decrease of diabetes rate from first cycle year to last cycle year, can I get the slope of changes? Then I want to compare two groups trend, such as male vs. female. Both trend from male group to female group were same or not?
the second similar question was related with continuous variable "age" when diabetes happened. Was there a trend of increase/decrease of average age when diabetes happened? if there was a trend of increase/decrease of age, how to compare male vs. female's age when diabetes happened?
Many thanks in advance for your looking at my questions, and provide some comments on it.