Each person would have 24,000-40,000 sites where there is a base substitution on one copy of their gene (hence heterozygous) where the substitution results in a change in amino acid (non-synonymous substitution).
I agree with the previous answer. The vast majority of these substitutions are neutral/non-pathological, contributing to the phenotypic diversity among people.
When you compare an individuals' coding region to the reference genome, you will find 24,000-40,000 heterozygous amino acid changes. The majority of these variants are expected to have NO dramatic effect on the function of the encoded protein.