Sounds spurious to me. I imagine people that take regular cold showers are also more likely to push themselves at work. To make this really interesting, you'd need to have a blind control group of people who are paid to have cold showers, and see whether their absence rates go down when they start.