continuous treatment variable: How to define treatment and control values

[HaibinLi0817]

Thank you very much for providing this valuable package.

I have a continuous treatment variable, namely BMI. How should I define the treatment and control values? For example, in the study (https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2777980), how should I set the values of a0 and a1 for HR for a 5-unit change? If a1=5 and a0=0, or a1=25 and a0=20, the results change. In addition, most studies did not report the values of a1 and a0, only showing the effect size. Are there specific rules for determining these a1 and a0 values?

In the example provided, a1=1 and a0=0 for a continuous variable. However, when I attempt to set different values for a1 and a0, the results change.

Best regards, Haibin Li