Most of what researchers know about everyday drinking comes from people remembering it. How many drinks. How it felt. When things got blurry. Researchers usually collect those memories the next morning, which is about the worst moment to ask.

A wrist sensor gives you another way in. The BACtrack Skyn reads alcohol through the skin and records the curve across the night. A team paired it with phone surveys for 11 days. I build a drinking app, so their question is one I care about: does the shape of that curve match how people feel while they're drinking?

Mostly, yes. The rest of the answer involves weekends.

What they did

Researchers recruited 50 moderate drinkers aged 21 to 25. Each one wore the Skyn for 11 days straight and went about normal life.

The "how it feels" side came from EMA (ecological momentary assessment), which just means short phone surveys answered in the moment. Participants filled one out after every drink. On nights without alcohol they still got several surveys, which gave the team a baseline. The surveys covered:

  • stimulation and sedation, on a brief scale built for alcohol's up-then-down effects
  • mood
  • craving
  • how intoxicated they felt

Every morning they also reported how many drinks they'd had the night before.

From the sensor, the team pulled three features of TAC (transdermal alcohol concentration, or alcohol measured through the skin):

  • Peak. The highest point of the night.
  • Area. The area under the curve, roughly the total alcohol exposure across the evening.
  • Rise. How fast the level climbed.

The raw sensor data went through a separate processing tool, the Skyn Data Manager.

What they found

First, the sensor passed the basic sanity check. All three TAC features went up with the number of drinks people reported the next morning, and the link was strong.

Then came the part I was curious about. The team used mixed models, which can handle many readings from the same person over many days. Here's how the sensor features lined up with what people reported in the moment:

  • Peak. A higher peak went with more stimulation, better mood, more craving, and feeling more intoxicated.
  • Area. Total exposure tracked the same four.
  • Rise. A faster climb predicted more stimulation, more craving, and feeling more intoxicated. Mood isn't on this list.

Sedation doesn't show up among the predicted outcomes in the abstract at all. I'd want the full numbers before reading anything into that.

The weekend effect is where it gets interesting. Peak TAC predicted intoxication more strongly on weekends than on weekdays. On a Saturday, a higher peak meant a bigger jump in how drunk people felt than the same difference did on a Tuesday.

The abstract doesn't say why. My guess: weekend nights run longer, happen in louder places, or simply get more of the drinker's attention. (Only guesses. Nothing in the study tests them.)

What it means

The speed result is the one I keep coming back to. Rate of rise tracked craving. When the curve climbed faster, people reported wanting more. The data is correlational, so the direction isn't settled. Maybe a fast climb feeds craving. Maybe people who already want a drink drink faster. Either way, speed and wanting showed up together, and that's worth knowing about yourself.

This is where AlcoBalance fits. It doesn't read your skin. It works from the drinks you log, with a calculation tuned to you, not averages. But it answers the same question the sensor does: how fast am I climbing right now? On a weekend night the pace can creep up without you noticing. See the climb while it's happening, and you have room to ease off on purpose. Before the peak, while you're still steering the evening.

The weekend effect points somewhere else. How drunk you feel depends on more than the alcohol in your body. Context shifts it. A study I covered earlier found that the most intoxicated people were the worst at estimating how drunk they were. Here the link between peak and feeling changed with the day of the week. The level and the feeling move together, but loosely.

One note on scope. These were 50 people aged 21 to 25 who drink moderately, so I wouldn't stretch the pattern to older or heavier drinkers. The abstract also doesn't mention checking the sensor against breath or blood readings. So the TAC numbers here tell you about drinks and feelings, but not your exact BAC (blood alcohol concentration).

My practical takeaway is small and specific. On weekend nights, how high you go seems to say more about how drunk you'll feel. Good reason to watch how fast you're climbing, not just how many drinks you've had.

Source: Alcohol, Clinical & Experimental Research, DOI