Does Affectionate Touch Help Everyone?

There is a person in a lot of relationships who does not much like being touched. Not during a fight, not on the sofa, not in passing in the kitchen. It is one of the more common quiet grievances a couple carries, and it usually gets explained the same way by both of them. He is just not a physical person. She needs less of that than I do.

A group of researchers took that explanation seriously enough to test it, and registered their prediction in advance.

What they did

The prediction was about attachment avoidance, which is the measured tendency to be uncomfortable with closeness and to prefer distance and self-reliance. People higher in avoidance report more negative attitudes toward touch. So the sensible hypothesis is that touch does less for them.

The team preregistered two hypotheses. First, that greater attachment avoidance would weaken the link between touch and wellbeing. Second, that avoidant people are less well partly because they get less touch in the first place.

They tested this across three studies, two of which were dyadic and included both partners. The first was a survey of 1,604 people. The second brought 80 couples into a laboratory and recorded three conversations, so touch could be observed rather than recalled. The third followed 98 couples through a 28-day daily diary.

Using recall, observation and daily reports in the same paper is a real strength, because each method fails differently.

What they found

The first hypothesis, the intuitive one, did not hold.

Touch was associated with greater wellbeing regardless of level of attachment avoidance. People who prefer distance were not getting less out of touch. They were getting about the same amount out of it as everybody else.

What was true is that avoidant people had less of it. Attachment avoidance was associated with lower wellbeing and with less frequent touch, and in most of the analyses the reduced touch statistically explained part of the gap in wellbeing. So the difference between the two groups sat in the amount, not in the benefit.

The authors’ own summary of this is worth stating plainly. It underscores the importance of touch even for people who value distance and autonomy.

We would flag that this is the kind of result that usually does not get published. The headline hypothesis failed. What makes the paper trustworthy is that it was written down in advance, so nobody could quietly turn a null moderation into a finding after the fact.

An honest reading of the limits

A null result is not proof of no difference. Failing to find that avoidance weakens the touch link is not the same as establishing that it never does. A smaller effect could be real and this design could miss it. The honest statement is that a preregistered test looked for the difference and did not find it.

It is still correlational, and the direction is genuinely unclear. Days with more touch were days with better wellbeing. People who feel better may reach out more, which would produce exactly this pattern with the causal arrow reversed. The laboratory observations help a little. They do not settle it.

The mediation is a statistical claim, not an experiment. Saying that less touch explains part of the wellbeing gap for avoidant people is a model fitted to correlational data. It held in the survey and in the daily diary, and it did not hold in the laboratory study, and the authors report it that way.

One control removes a good deal of it. In the daily diary study, once the authors accounted for how good the relationship was overall, the links between avoidance and mood, and between avoidance and touch, stopped being significant, and so did the mediation. That is their own reported check and it belongs here. It suggests some of what looks like an effect of a person’s attachment style may be an effect of the state of the relationship they are in.

A later paper from the same first author found a cultural limit. A 2025 study comparing Latin American and Swiss samples reports that the frequency of affectionate touch is not always associated with wellbeing, and that the touch measure itself did not behave equivalently enough across those groups to compare averages. The person best placed to find the boundary of this finding went looking and found one.

Samples are Western and self-selected. North American and largely white, with the laboratory couples averaging 23 years old and about two years together. Couples who agree to be observed touching each other in a laboratory are not a random draw from the population, and there is an obvious direction to that bias in a study about touch. That laboratory sample also had no power analysis behind it, because the data had been collected for a different purpose.

Nothing here measures a person being touched more than they want. The studies looked at how much touch happened and how people were doing. They did not test what happens when one partner increases touch on purpose because they read that it is good for the other person. That is a different question and this paper cannot answer it.

What this means for your fight about closeness

The fight this speaks to is usually not about sex. It is about the smaller currency: a hand on the back while passing behind a chair, sitting close rather than across, whether you are touched at all on an ordinary Tuesday.

What this research changes is the story each of you tells about the other. The common version is that one of you needs touch and the other does not, and that the gap is a difference in wiring that has to be lived with. On this evidence that story is probably wrong in an important way. The person who reaches out less is not getting less out of it when it happens.

That reframes the conversation from a mismatch in needs to a difference in approach, which is a much more workable thing to talk about. It also suggests the more distant partner has more at stake in the outcome than either of you has been assuming.

What it does not license is pressure. A finding that touch is good on average is not permission to touch a person who has said they do not want to be touched right now, and a partner who starts making contact because a study said it helps will be noticed doing it. The result is about what happens when touch is welcome. It says nothing at all about touch that is not.

Where a session with Elena fits

Elena is an AI coach, and she works with what you each say rather than with anything she can observe. She cannot see how you sit on the sofa.

What she does with this is narrow. When one of you describes the other as not a physical person, she asks the other one directly, in the private room, whether that is how it feels from the inside. The answers frequently differ, and that difference is usually more useful than another round of the same conversation.

She does not set targets for physical affection, and she will not suggest one. This is an area where a coach handing out a number would be doing something the research does not support.