Groups of rational numbers and remarks on the analytic part of Hooley's proof