The latest Pet Shop Boys album has a song rendered in full histrionic castrato with the words 'Love is a catastrophe, look what it's done to me'. Although Neil Tennant probably did not have conventional heterosexual marriages in mind, the words apply to all too many of them. And if hell is other people, so is divorcing them. While it may be a relief for some, overall, divorcees are at least three times more likely to be admitted to a mental hospital than persons in intact marriages. Divorcees also suffer much higher rates of physical illness, alcoholism and suicide.

Conventional thinking sees this catalogue of misery as primarily the result of the disharmonious gender rancour leading up to the divorce and the lonely penury which often follows it. But that is only part of the story. It may also be because divorcees are liable to be very unhappy, whoever they are with.

Couples with one partner who is depressive are more disharmonious. When 56 married depressives were followed for a two-year period, they were nine times more likely to divorce than the general population. But which comes first, the marital disharmony or the depression?

Until the publication of Myrna Weissman's study The Depressed Woman in 1974, it was generally assumed that the depressed were silent, compliant and passive in all contexts. But this was based on merely interviewing them once, rather than observing them at home.

She found they were paranoid and aggressive to close intimates, especially partners and children, and the same has since been shown to be true of depressed men.

Partners of depressives are themselves more likely to be depressed, to get physically ill, to abuse alcohol and to commit both suicide and homicide. Compared with couples where neither is depressed, in couples with one depressive, the depressed partner is more likely to be inflexibly overbearing in solving disputes.

The couple feel miserable about their relationship, are secretive and uncommunicative, providing little support to each other. In short, depressed people tend to be hell to live with as well as living in hell.

The implications are intriguing. Rates of depression among 25-year-olds have risen between three and 10 times compared with 1950; rates of divorce are seven times higher.

There can be little doubt that the increased depression will have increased the divorce rate. On top of that, the greater ease of divorce will have enabled more depressed people to 'act out' (avoiding feelings through action) their hostility and paranoia towards their partner by divorcing them rather than getting treatment for it. The consequent increased number of divorcees - since divorced status increases depression - will have then further boosted the depression stats.

However, before drawing these conclusions, there is an important caveat. It could be that marriage is depressing, that depressed marital partners become so as a result of being with the wrong partner, or just by being in an 'enslaving' institution, rather than that the problem is 'in' the depressed person... next week's subject.