A few years ago, in a clinical trial my colleagues and I ran, something happened that I still find myself thinking about. We asked postmenopausal women who were having frequent, disruptive hot flashes to change one thing: their diet. They shifted to a low-fat, plant-based pattern with a serving of soybeans each day. We weren’t testing a drug or a hormone. We were testing food.
After 12 weeks, moderate-to-severe hot flashes fell by 88 percent. Half the women stopped having them altogether.
I want to be careful here, because a result like that invites overreach. This was one trial, and diet won’t do this for every patient. But the size of the effect is worth sitting with, not because it proves food should replace hormone therapy, which remains genuinely effective and appropriate for many women, but because of why it happened. The most honest answer is that we don’t fully know. That uncertainty is the point, and I think it points at a blind spot in how we practice.
For a century, medicine has made extraordinary progress by breaking the body into parts and assigning each to a specialist. Cardiology takes the heart. Endocrinology takes the hormones. GI takes the gut. Neurology takes the brain. This division of labor gave us antibiotics, insulin, statins: some of the most effective interventions in human history. It works because it isolates a problem, finds the broken component, and fixes it.
Real physician voices, twice a week
Free, and one click to unsubscribe.
We explain hot flashes the same way in every exam room: Estrogen declines, so the thermostat goes haywire. Replace the missing estrogen, and symptoms ease. Clean story, single cause, single fix.
But biology doesn’t always respect our organizational chart. Vasomotor symptoms track with body weight, vascular function, core temperature regulation, inflammation and (increasingly, the evidence suggests) the gut microbiome. None of these systems runs on its own. They’re wired together.
A dietary change touches all of them at once. Fiber alters hormone handling. Soy compounds interact gently with estrogen receptors. Weight loss changes thermoregulation. Diet reshapes the microbiome, which feeds back on metabolism and hormone signaling. We didn’t pull one lever in our trial. We nudged a whole network in a healthier direction at once. The dramatic result may simply be what happens when a lot of small pushes move together.
That’s the blind spot, and it’s a design problem, not an intelligence problem. Our entire research and treatment apparatus is built to isolate one variable at a time. The randomized trial, our best evidence tool, is engineered to hold everything constant except a single intervention so we can see its effect cleanly. Regulators approve drugs against defined molecular targets. We organize ourselves by organ. This machinery is superb at finding single causes with single fixes, and far less suited to interventions that work by touching everything at once.
Diet, activity, sleep, stress: These act across many pathways simultaneously, spilling across the boundaries between our specialties. That makes them harder to study, harder to attribute credit to, and easier to underrate on rounds and in guidelines. We’ve built a system that sees single targets in sharp focus and treats networks as noise.
Menopause is just the example I happen to have trial data for. It’s a window into a bigger pattern. Obesity involves metabolism, sleep, stress and environment together. Type 2 diabetes plays out across liver, muscle, pancreas, brain and gut simultaneously. Cardiovascular disease emerges from inflammation, metabolism, vasculature and genetics interacting over decades. As systems biology and microbiome science mature, our patients look less like a set of separate organs and more like a network, where a change at any node ripples through the rest.
The timing matters for those of us in obesity medicine, diabetes care and menopause treatment specifically, because our fields are being reshaped right now by a new generation of highly targeted drugs, GLP-1s foremost among them. These therapies can be genuinely transformative, and this isn’t an argument against prescribing them. It’s an argument for humility about what a single target can and cannot do on its own, and for not letting a networked problem get treated as if it had one cause simply because that’s the shape of the tool in hand. Real benefit, the kind our menopause participants experienced, is easy to leave on the table when it doesn’t fit a billing code or a single-mechanism drug trial.
The harder task ahead isn’t just finding better targets. It’s building research, and clinical practice, that can capture what happens when many systems shift together instead of treating that as noise to control away. The body has been operating as a network all along. It’s our science, and our specialty structure, that are only beginning to catch up.
Hana Kahleova is an endocrinologist.
