Every dad joke about the "dad bod" assumes the same thing: he's just eating more. Sitting on the couch instead of the sidelines. Letting himself go along with the schedule.

Nobody in that joke considers the possibility that something in his body changed on its own, on a script he didn't write, for reasons that have nothing to do with willpower.

But it did. Before the baby ever arrives.

A team of anthropologists spent four and a half years tracking testosterone in 624 men in the Philippines. At the start, the men were single, and the ones with the highest morning testosterone were the most likely to go on to become partnered fathers — high T, it turns out, is good for finding a mate. But once they became fathers, something flipped. Testosterone dropped by roughly a quarter, on average. And the men who threw themselves hardest into hands-on caregiving saw the steepest declines (Gettler et al., 2011).

That's a real, measurable effect — but it comes with a catch worth sitting with. The same men whose testosterone dropped the most were also the men who chose to be the most involved. It's possible fatherhood is doing something to their biology. It's also possible the kind of man who already runs a little lower on testosterone is simply more inclined to lean into the caregiving in the first place. The study can't fully separate the two, and neither can we.

What makes the pattern more interesting is that it isn't fixed — it bends around whatever a given culture actually asks of its fathers. Among the BaYaka, a forager community in the Congo Basin, the men most respected for sharing food and looking after children had the lowest testosterone of the group. Compare that to their neighbors, the Bondongo, a fishing and farming society, where being a strong provider maps onto a slightly different hormonal profile entirely (Gettler et al., 2020). Same biological system, different cultural instructions for what to do with it.

There's also a hormone that rarely gets mentioned outside a delivery room: prolactin. In one of the earlier studies to track expectant fathers directly — thirty-four couples in Canadian prenatal classes — prolactin rose in the weeks before birth, while testosterone and estrogen both dipped after. The men who showed the most pregnancy-adjacent symptoms of their own — fatigue, weight gain, the whole "sympathetic pregnancy" cluster — were also the ones who responded most strongly to cues from the infant (Storey et al., 2000). The body appears to be laying groundwork before there's even a job to do yet.

And it isn't just hormones. A longitudinal brain-imaging study out of USC scanned 38 first-time fathers twice — once during their partner's pregnancy, once around six to eight months after the birth. Their cortical gray matter measurably shrank. But the men who lost the most volume don't sort neatly into a "good outcome" story. The steepest drops showed up in the men most bonded to the pregnancy beforehand and most engaged with their infant afterward — and, separately, in the men reporting the worst postpartum sleep and the highest depression, anxiety, and distress scores. The same measurable change tracked both engagement and strain at once (Saxbe & Martínez-García, 2024).

Here's the part I keep turning over. A companion study from the same lab, run independently on two continents — one sample in Spain, one in California — found the same pattern both times: the volume loss was concentrated in the cortex. The older, more automatic subcortical structures were preserved, in both samples (Martínez-García et al., 2022).

I want to be careful here, because it would be easy to oversell this. The study didn't scan mothers and run a head-to-head comparison — so I can't tell you fatherhood produces "half" of whatever happens in mothers, and I'm not going to claim it does. What the data does show, replicated twice, is a location: the change in fathers shows up in the newer parts of cortex tied to social cognition and attention, and it leaves the older, automatic machinery alone. That lines up with the hormone data — a change that's built in the part of the brain responsible for effortful, flexible attention, not the part responsible for automatic response.

Steve Stewart-Williams has an explanation for why any of this would be conditional in the first place, and it's worth laying out plainly. His argument: parental investment, in evolutionary terms, is a resource an organism allocates, and the allocation should track how confident the investor can be that the investment is actually theirs. Mothers carry paternity certainty built into biology — the child came out of her body. Fathers, in his account, never had that certainty, which is why he argues paternal care tends to run more conditional, more responsive to context, than maternal care across every species that provides it at all. In his framing, there was no pressure to build fathers a hair-trigger nurturing instinct — only pressure to build something quieter, that switches on with proximity and evidence rather than firing automatically at birth.

This seems worth sitting with rather than settling. I find the "why" harder to buy than the "what" — it's a tidy story, and it might even be right, but it's also the kind of story that's difficult to falsify. Almost any pattern of paternal behavior could probably be retrofitted to "the mechanism favored conditional investment." Read the same four studies without his frame, and you land somewhere just as solid: proximity triggers the biological response, full stop. That part doesn't seem to need a theory of why it evolved to be true. The data holds up on its own — I'm just not sure the story wrapped around it is carrying the weight it claims to.

What isn't in question, whichever explanation you find convincing, is what "involvement" is actually doing. It isn't just a moral choice layered on top of indifferent biology. Proximity itself appears to be the trigger — the contact, the caregiving hours, the smell of the kid — that pulls the hormonal and neural system into gear at all. The capacity for this kind of change in fathers doesn't arrive pre-loaded and waiting; it has to be activated by showing up, and showing up is squarely a matter of individual choice. The biology is real. It just isn't self-executing the way it is for a mother carrying and delivering the child.

Richard Reeves has spent a career documenting the structural side of this — divorce, custody defaults, workplaces that assume fathers won't take leave — and none of that is a biology problem. It's worth noticing, though, that if the paternal system really does require proximity to switch on, then every structural force that keeps fathers at a distance isn't just a social cost. It may be quietly preventing a biological process from ever getting the chance to run.

I don't think that settles anything about how much of modern fatherhood is nature and how much is circumstance. It just makes me wonder whether the two questions were ever really separable to begin with.