I fitted a model with Proc Surveylogistic. There are categorical and continuous independent variables. The ManN10 is one of the continuous variables. The fitted model Type 3 analysis indicates that the P value for ManN10 is 0.01, but the odds ratio for ManN10 is 1 and the interval is 1-1 as well. How do we explain it? Odds ratio of 1 is supposed to mean that the odds is equal, but why type 3 analysis gives a significant result?