In my opinion, it's much cleaner and easier to understand if you extract the retrieval of environment variables into a utility (you probably want to include a check to fail fast if an environment variable is not set anyway), and then you can just mock the utility.
// util.js
exports.getEnv = (key) => {
const value = process.env[key];
if (value === undefined) {
throw new Error(`Missing required environment variable ${key}`);
}
return value;
};
// app.test.js
const util = require('./util');
jest.mock('./util');
util.getEnv.mockImplementation(key => `fake-${key}`);
test('test', () => {...});