Were there any realistic way to check the percentages, I would be willing to make a substantial bet; I don't think you would find anyting close to a significant correlation with belief or unbelief. I don't think that sickness like that is any respector of nominal beliefs.