It's strange, but recently all I seem to hear about are young scientists whose relationships have broken down. OK, that's a slight exaggeration. But I did run into one of Naturejobs' former writers for Graduate Journal at a postdoc retreat the other week, only to find out that they had just undergone a divorce. Not long after, a chat with a friend revealed that her brother's relationship had ended when he began his PhD studies. These tales of woe reminded me of another Graduate Journal writer who got divorced while he was working on his dissertation. And that got me thinking about the effect that building a scientific career can have on relationships.

I don't have any facts or figures about relationship issues among PhD scientists, but I suspect that they get married later than the average, and experience more romantic difficulties than normal earlier in their career. But why? First, I suspect that location can exert a significant pressure. When someone selects a postdoc or a first job, it's unlikely to be close to wherever a spouse or a partner already works — especially if the partner is also in academia. That means young scientists must either limit their job search, expect their partner to move with them, or carry out a long-distance relationship. A choice that is less than satisfactory.

Then there's time. Running experiments, teaching and writing papers and grants all consume more time than a 'normal' working week. It takes a very understanding partner to put up with that day after day. No one likes to be stood up for a romantic dinner because their partner had to run to the lab to check on gels.

Finally, there's stress, which tends to come from things that are beyond your control. Even if you crank out first-rate data, draft brilliant papers and write compelling grant applications, ultimate success often lies in someone else's hands. It is no mean feat keeping your composure in the face of such ceaseless challenges.

I have no solid data to support this thesis, other than a few anecdotes and some loose correlations. I'd love to be proved wrong, but need some better data. So please send any relevant stories or figures to me at