Already created First names, last names and gender (gender is in accordance with the first name). 
We downloaded a huge csv file where a company gave names and gender of loads of people.

For height: look at distributions between boys and girls and take means and variances for each and sample from normal distributions for each gender

BMI and height :  basically the same

Country and city : do the same as for first name and gender, we found such a file

Case control status : sample from binomial distribution that has a prob of 0.5

Education level : 5 different levels, we sample from uniform distribution

Gene expression and SNP values : we have 10 different gene expression features and 5 different SNP features. Each gene can take any number between -1 and 1. Each SNP can take any value between (0,1,2). So do the same as before, that is sample from uniform distribution.