Why am I always cold when everyone else is fine?

30 July 2026 · 11 min read

If you're always cold when everyone else in the room is fine, you're not imagining it and you're not being difficult. The temperature that feels right to you is a real number, and it can sit as much as 10°F (6°C) away from the number that feels right to the person at the next desk. Same room, same air, two correct answers.

I run cold. That's the only reason this post exists.

Comfie started out dressing everybody the same, off the same forecast, and for about a third of the year it was quietly wrong for me. Not dramatically wrong. One layer short, often enough that I stopped trusting it. My first assumption was that the weather data was bad. It wasn't. The data was fine and I was the variable, which meant the fix had to live in the app rather than in the forecast.

So I went and put a number on it. Here's what that number turned out to be, where it comes from in your body, and the honest bit about how much of it you can change.

Your comfortable temperature is a number, and it isn't the average one

Some people are comfortable several degrees away from the average, because body size, muscle mass, circulation and sex all change how much heat you produce and how fast you lose it. The usual gap is 4 to 10°F (2 to 6°C). It is physical, it is measurable, and most of the time it isn't a medical problem.

Comfie handles that gap as an offset, applied in degrees before any clothing gets chosen. Setup asks one question: do you run cold, about average, or hot. Answer "I run cold" and every forecast gets 4°F (2°C) subtracted before the outfit is picked. A 62°F (17°C) morning is dressed as a 58°F (14°C) morning. Answer "I run hot" and it moves the other way by the same amount.

Then it watches. Every day you rate moves your number by half a degree Fahrenheit, in the direction the rating points. Say a day was too cold and tomorrow gets dressed half a degree warmer. Say it was just right and nothing moves at all, which is the part that took the longest to get right: the correction has to hold once it's found, not sag back toward the average the moment you stop complaining. The running total stops at 6°F (3°C) either way, so a bad week can't send your profile somewhere strange.

Four degrees declared plus six degrees learned is ten. Set against someone sitting at the opposite extreme, the distance between your two correct answers is 20°F (11°C). That is not a preference. That's most of a season.

What a few degrees actually changes

Degrees are abstract. Outfits aren't. So here is one real morning, 48°F (9°C) on the feels-like scale rather than the raw air temperature, run through Comfie's engine four times with nothing changed except who's wearing it.

Swipe the table sideways for the rest of each row.

Top row and bottom row are the same weather. One of them is a hoodie. The other is a puffer jacket, boots and a hat.

That's the thing worth taking away, and it's why "just check the temperature" was never going to be enough. Two people can read the identical forecast, follow it correctly, and need different clothes.

Want the version for today instead of for 48°F? The live demo runs your city through the same engine, no download.

The same garment, three different temperatures

Another way to see it. A hoodie on its own, with nothing over it, is the single most common outfit decision there is. Here's the window where Comfie reaches for one, drawn for each profile.

A 60°F (16°C) day is a hoodie day for an average person and a long-sleeve day for someone who runs hot. Push the runs-cold profile to its fully learned extreme and the hoodie window climbs to 62 to 72°F (17 to 22°C), which is a temperature most people call pleasant. The rest of the ranges behind that live in hoodie and jeans weather.

Why do I get cold easier than others?

Five things, roughly in order of how much they matter for a healthy adult.

• You're smaller. Heat is made in your volume and lost through your skin. Shrink a body and the surface shrinks slower than the volume does, so a small person carries more skin per unit of heat-producing mass and sheds warmth faster doing nothing at all. Children are the extreme case, which is why they get cold quickly and warm up quickly.
• You have less muscle. Muscle burns energy even at rest and is most of what your resting metabolism actually is. More muscle means a warmer body idling. This is also why the same person feels colder after a few sedentary years, at the same weight.
• Your circulation sacrifices the edges. When your body decides to conserve heat, it narrows the blood vessels in your hands and feet and pulls warmth inward to protect your organs. That's the system working correctly. It just means fingers and toes get abandoned first, and they're where you notice cold.
• You're female. Resting metabolic rate runs around 23% higher in men, so a man sitting still generates more heat than a woman sitting still. Women's hands tend to sit about 1.5°C cooler, and estrogen both lowers body temperature and slows blood flow to the extremities. None of this is small, and all of it compounds in a room set for someone else.
• Something medical, sometimes. An underactive thyroid slows your metabolism and your heat production with it. Iron deficiency leaves your blood carrying less oxygen. Raynaud's makes fingers and toes clamp down hard in response to mild cold. These are common and testable.

Some of it is only what you're used to

Now the part that cuts against everything above.

Bodies adapt to cold, and faster than most people expect. Seven days of regular cold exposure has been shown to cut shivering intensity by roughly 20%. Four weeks raised brown fat volume by 45% in one study, brown fat being the tissue that burns energy to make heat instead of storing it.

Here's what stops me overselling that. In at least one of those studies, the time it took people to start shivering didn't move at all. Their bodies got better at producing heat quietly, without the shaking. They didn't stop feeling cold.

Which matches my own experience closely enough to be annoying. Two winters of walking to work made me better at being cold. It did not make me warm.

The practical version: if you moved somewhere colder in the last year, give it a full season before you decide you run cold, because you're probably still adjusting. If you've been in the same climate for a decade and you're still the one reaching for a cardigan in June, that isn't acclimatization. That's you, and you may as well dress for it.

Why am I always cold in the office?

This one has a specific and slightly infuriating answer.

The formula underneath most building air conditioning was developed in the 1960s, and it has a reference human baked into it: a 40-year-old man weighing about 154 pounds (70 kg). Every thermostat setting derived from it is answering the question "what would that guy find comfortable."

In 2015, Boris Kingma and Wouter van Marken Lichtenbelt at Maastricht University measured what that assumption costs. Writing in Nature Climate Change, they found the standard may overestimate female metabolic rate by up to 35%. The women they measured were running 20 to 32% below the value the formula assumes. Their sample was 16 young women, which the authors flagged themselves as small, so read it as a strong signal rather than a settled constant.

Four years later Tom Chang and Agne Kajackaite put 543 people through maths and verbal tests in a room they moved between 61 and 91°F (16 and 33°C). Women scored better as the room warmed. Men scored slightly worse. The women's improvement was the larger and more clearly measured of the two effects, which is a real finding about productivity and not just comfort. It ran in PLOS ONE in 2019.

Then stack the ordinary stuff on top. You sit still for seven hours, and sitting still generates almost no heat. The vent is above your desk and not above anyone else's. You can't add a blanket without it becoming a whole thing. None of that appears in the formula either.

So the office isn't calibrated to you. It's calibrated to a 1960s average, and you're the one absorbing the difference.

How to find your own number

You don't need an app for this. Two weeks of paying attention will do it.

1. Read the feels-like number, not the temperature. The headline figure is air measured in the shade. Wind, humidity and sun move what your skin actually gets by a lot, so start from the number that accounts for them. Every comparison below falls apart if you begin from the wrong one.
2. Log the day in three buckets. When you get dressed, write down the feels-like figure. At the end of the day mark it cold, right, or warm. Nothing more elaborate than that.
3. Find where your "right" days cluster. After ten or so entries the pattern shows up. If your comfortable days sit consistently four or five degrees below where the usual advice puts them, that's your offset, and you now know it.
4. Apply it by hand from then on. Subtract your number from every forecast before you decide what to wear. Someone who runs 5°F (3°C) cold should treat a 60°F day as a 55°F day, permanently.

Two details that make the result better. Dress your hands and feet ahead of your torso, because the extremities get shut down first and they're where the feeling of being cold actually starts. And eat something before you go out in the cold, since a body short on fuel throttles heat production to conserve what's left.

What Comfie does with your number

The app is that same loop, without the notebook.

You answer the run-hot-or-cold question once, which sets the starting offset at plus or minus 4°F (2°C). After that, each day you rate moves it half a degree, capped at 6°F (3°C) of learned correction in either direction. Consistent feedback converges on your real number and then holds there.

Two design decisions in that loop are worth spelling out, because both came from getting it wrong first.

It keeps three separate offsets, not one. I built it with a single number and it didn't work, because running cold in February tells you very little about how you handle August. So there's one offset for cold days below 55°F (13°C), one for mild days in between, and one for hot days above 72°F (22°C), each learned on its own. Plenty of people are a layer short all winter and completely fine in summer, and one number can't express that.

And it ignores the days it can't do anything about. When it's 90°F (32°C) and you're already in the lightest thing the app owns, a "too hot" rating isn't feedback about the outfit. It's a complaint about the weather, and it's fair. But if the learner took it seriously, a single heat wave would drag your whole hot-weather profile lighter and you'd get handed shorts on the next 74°F (23°C) day. So those ratings are recorded and skipped. Same rule at the freezing end.

The web demo on this site deliberately runs without any of this. It's the forecast alone, dressed for an average person, which is the honest way to show a stranger what the engine does. The personal part needs to know you, and that lives in the app.

Quick answers

Why am I always cold but my body temperature is normal?

Because feeling cold is driven by your skin temperature, not your core. Your body defends the core first and gives up on the surface, so your hands and feet can be genuinely cold while a thermometer under your tongue reads a perfectly normal 98.6°F (37°C). A normal reading rules out a fever. It doesn't rule out being cold.

Is being cold all the time a sign of something serious?

Usually not on its own. Cold sensitivity that has been steady your whole adult life is almost always just your body. What's worth a doctor's appointment is a change: getting noticeably colder over a few months, or cold plus fatigue, weakness, hair thinning, unexplained weight change, or fingers that turn white and then blue. Those point at thyroid, iron or circulation, and a blood test settles it quickly.

Why am I always cold in the office but fine at home?

Three things stack up. Office air conditioning is set by a formula built in the 1960s around a 40-year-old man weighing about 154 pounds (70 kg). You sit still for hours, which produces almost no heat. And you can't put on a blanket. At home you move around, dress how you like, and set the thermostat yourself, so the same body feels completely different.

Can you train yourself to stop feeling cold?

Partly. Regular cold exposure measurably changes your body: one study cut shivering intensity by about 20% in seven days, and four weeks raised brown fat volume by 45%. But in at least one of those studies the time it took people to start shivering didn't change at all. You get better at handling cold. You don't stop noticing it.

Why are women colder than men in the same room?

Resting metabolic rate runs roughly 23% higher in men, so a man sitting still is generating more heat than a woman sitting still. Women's hands also tend to sit about 1.5°C cooler, and estrogen slows blood flow to the hands and feet. Add a thermostat calibrated to male metabolism and the same room is genuinely two different rooms.

You're allowed to just dress differently

The reason this bothers people isn't really the temperature. It's the implication that everyone else got the memo and you didn't, so you must be doing something wrong.

You aren't. The average is a construct. Nobody is the average, the office thermostat is answering a question about a man from 1965, and the gap between you and the person beside you is a measurable number of degrees rather than a character flaw.

Find your number. Then wear the extra layer without explaining yourself. And if the layer in question is the one that goes over the hoodie, that's a whole separate decision.

Let Comfie learn your number

The live demo on the homepage dresses an average person from real weather. Type your city and see it work, no download.

The iPhone app is the one that matters here. It asks whether you run hot or cold, then adjusts from your ratings until a 55°F (13°C) day means for you what it actually means. Free, no ads.

Where these numbers come from: the offsets, caps and learning zones are the real values inside Comfie's comfort profile, and every outfit above was produced by running the engine at that temperature rather than estimated. Weather data comes from Open-Meteo. Temperatures are feels-like, not raw air temperature, and Celsius figures are rounded to the nearest degree. The physiology is cited inline; none of it is medical advice.