After an Affair: What Five Years Showed

There is very little good data on what happens to couples after an affair. Most of what gets written is clinical experience, which is worth something, or personal essay, which is worth something else.

This study is one of the few places where couples who had been through an affair were followed for five years with the same measures used on everyone else. It is a small study and it says something that most writing on the subject does not.

What the study did

Between 1997 and 2001, 134 seriously distressed married couples enrolled in a randomised trial of couple therapy at two American universities. They were assigned to one of two well-established approaches and treated for around eight months. Nobody was selected for or against having had an affair.

Afterwards, 19 of those couples were identified as infidelity couples. They split into two groups that turned out to matter enormously:

• Revealed, 14 couples. Six disclosed the affair before therapy started; eight disclosed during it.
• Secret, 5 couples. The affair only came to light after treatment had finished, which means therapy had been conducted without it.

The remaining 115 couples had no known infidelity. Everyone was reassessed roughly every six months for five years after treatment ended.

An earlier paper on the same sample had already reported something hopeful: infidelity couples arrived more distressed than everyone else, and those who disclosed improved more over the course of therapy than the non-infidelity couples did. The five-year follow-up asked whether that held.

What it found

Partly. And the part that did not hold is the part worth reading carefully.

More infidelity couples divorced. Across the five years, 53% of the infidelity couples divorced against 23% of the others, with the odds of divorce about 3.7 times higher. The confidence interval on that ran from 1.4 to 9.7, which is a wide range and an honest reflection of how few couples this is.

Concealment separated the two infidelity groups. Four of the five secret couples divorced. Six of the fourteen revealed couples did.

And now the finding that almost never gets quoted. Among the couples who stayed married, infidelity made no detectable difference. Satisfaction was statistically indistinguishable between infidelity and non-infidelity couples who remained together, and their marital stability was too. Those couples also reported satisfaction increasing over the five years, regardless of whether there had been an affair.

The authors describe this as two pathways. Some couples who had been through an affair carried on improving and became indistinguishable from couples who never had one. Others deteriorated and separated. There was not much middle.

Which treatment a couple received made no difference to any of this.

What the study cannot tell you

Nineteen couples. That is the whole infidelity sample, and it splits into fourteen and five. The authors say directly that with groups this size, their estimates lack precision. The 80% figure describes five couples. Anyone quoting it as a law about secrecy is doing something the paper explicitly warns against.

Nobody assigned couples to conceal. Couples in whom an affair stayed hidden differ from couples who disclosed, in ways nobody measured. There is also something specific and strange about the secret group: their therapy was conducted while one partner was withholding the central fact of the marriage. It is not possible to separate the effect of secrecy from the effect of eight months of treatment aimed at the wrong problem.

This was a secondary analysis. The trial was designed to compare two therapies, not to study infidelity. The infidelity groups emerged from the sample rather than being recruited.

The comparison group is not clean. Some proportion of the 115 non-infidelity couples had affairs that were never disclosed to anyone, including the researchers. That would pull the two groups toward each other and make the real contrast harder to read, in an unknown direction.

These couples were all in therapy. Every one of them sought and completed around eight months of treatment with a trained clinician. Nothing here describes what happens to couples after an affair without that.

It is a specific time and place. Two American university clinics, couples recruited between 1997 and 2001, seriously distressed at intake.

And the good news is conditional. The finding that infidelity couples became indistinguishable applies to the couples who stayed married. It is a statement about survivors. Half of the infidelity couples were not in that group.

What this means if it has happened to you

Three things, stated as carefully as the data allows.

The first is that recovery is a real outcome and not a consolation prize. Couples in this study who stayed together after an affair were, five years on, doing as well as couples who had never had one, and getting better rather than worse. That is a genuinely different message from the one this subject usually carries.

The second is that disclosure sits on the better path. Every couple in the revealed group had the affair on the table where it could be worked on. Every couple in the secret group spent their treatment with it hidden. The numbers underneath that comparison are tiny and the direction is consistent with the whole clinical literature.

The third is the sober one. Even among couples who disclosed and did eight months of therapy, more than four in ten divorced within five years. This study does not support the idea that affairs are straightforwardly survivable given enough effort, and it would be dishonest to sell it that way.

Where a session comes into this

Here is the part where a product like ours is supposed to tell you it can help.

For this specific thing, the honest answer is mostly no. Every couple in this study had months of work with a trained human clinician, and even then the outcomes split hard. An AI coach is not a substitute for that, and affair recovery is close to the clearest case on the list. If an affair has happened in your relationship and you both want to try, a human couples therapist is what the evidence describes.

What Elena can honestly do is smaller and sits earlier. If there is something you need to say and every version you have rehearsed goes badly, The First Draft is ten minutes on your own, working on the words until they are sayable. Some people use it before a conversation they have been putting off for months.

That is a real thing and it is not treatment. This study is a reasonable place to be told the difference plainly.