Mongodb - aggregate by embedded document id overwrite the outer document _id

I have this kind of 'comment' model:

{ _id: <comment-id>,
  user: {
    id: { type: Schema.ObjectId, ref: 'User', required: true },
    name: String
  },
  sharedToUsers:          [{ type: Schema.ObjectId, ref: 'User' }],
  repliedToUsers:         [{ type: Schema.ObjectId, ref: 'User' }],
}

And I want to query for all comments which pass the following conditions:

1. sharedToUsers array is empty


2. repliedToUsers array is empty

But also, I want the result to contain only 1 comment (the latest comment) per user by the user id.

I've tried to create this aggregate (Node.js, mongoose):

     Comment.aggregate(
        { $match: { "sharedToUsers": [], "repliedToUsers": [] } },
        {
            $group: {
                _id: "$user.id",
                user: { $first: "$user" },
            }
        },
        function (err, result) {
            console.log(result);
            if (!err) {
                res.send(result);
            } else {
                res.status(500).send({err: err});
            }
        });

It is actually working, but the serious problem is that the results comments _id field is been overwritten by the nested user _id.

How can I keep the aggregate working but not to overwrite the original comment _id field?

Thanks