Sleep doesn’t happen in a vacuum. What you do during the day — and the hours before bed — shapes what your body does overnight. Here’s what the research says, and how Root helps you find your own connections.
The Problem With General Advice
Sleep hygiene advice is everywhere: avoid caffeine after 2 PM, stop alcohol three hours before bed, exercise in the morning. Some of this is backed by solid population-level research. But people vary enormously. One person’s evening espresso has no measurable effect on their HRV; another person’s afternoon coffee tanks their deep sleep.
Root’s Patterns tab is built around a simple premise: measure your response, not the average person’s.
How Tagging Works
Each night on the Night screen you can tag what happened that day. Root includes eight built-in tags:
| Tag | What it captures |
|---|---|
| ☕ Caffeine | Coffee, tea, energy drinks in the hours before bed |
| 🍷 Alcohol | Any alcohol within a few hours of sleep |
| 🏋 Exercise | Significant physical activity that day |
| 🌙 Late Night | Going to bed noticeably later than your norm |
| 😰 Stress | High mental or emotional load |
| 🤒 Sick | Illness or physical discomfort |
| ✈️ Travel | Time zone shifts or unfamiliar sleep environments |
| 🏠 Poor Sleep Environment | Noise, temperature, light, or partner disruption |
Tags take under five seconds to set. Over 30 days they become your personal dataset. You can also create custom tags in Settings to track anything specific to your own routines.
What the Research Says (and What Root Measures)
Caffeine
Caffeine’s half-life is 5–7 hours in most adults. A 3 PM double espresso still has roughly 50% of its caffeine circulating at 9 PM. At the population level, caffeine suppresses deep sleep even when you feel no trouble falling asleep.
In Root: after a few weeks of tagging, check your Patterns tab to see whether your restorative sleep or HRV is measurably lower on caffeine days versus non-caffeine days. The result is often surprising — both directions.
Alcohol
Alcohol is sedating in the first half of sleep (you fall asleep faster) but disruptive in the second half: it suppresses REM sleep, elevates heart rate, and fragments sleep cycles. Even one to two drinks typically shows up as elevated sleeping heart rate and reduced restorative sleep. The effect is often visible in Root data the morning after the very first tagged night.
Exercise
The relationship is positive but timing-sensitive. Vigorous exercise can temporarily elevate resting HR and lower HRV in the 24 hours post-workout as your body repairs muscle — this is normal and healthy. Regular moderate exercise consistently raises HRV over weeks.
Root’s Patterns tab shows whether your exercise days correlate with higher or lower next-night HRV. The lag is typically one night, which is the right comparison to make.
Screen Time and Late Nights
Blue light and stimulating content suppress melatonin and delay sleep onset. The Late Night tag captures the behavioural pattern (going to bed late) rather than screen time directly. If your late-night-tagged entries cluster with lower total sleep or later wake times, that’s your personal data showing the pattern in your own numbers.
Reading the Patterns Tab
The Patterns tab shows every tag you’ve used in the past 30 days, alongside an impact indicator (1–5 dots):
- Dot count — how much your sleep metrics differ on nights associated with that tag versus nights without it. Five filled dots means the tag correlates with a ≥28% difference on at least one metric.
- Dot colour — green means the tag correlates with better sleep metrics (exercise improving HRV). Clay/red means worse (alcohol elevating sleeping HR).
Root requires a minimum of 3 tagged nights before computing a correlation — this prevents a single outlier from creating a misleading pattern. The significance threshold is a 5% difference between tagged and untagged nights; smaller differences are filtered out as noise.
Tapping any tag opens a detail view showing exactly which metric is most affected and by how much — so you can see “on alcohol nights, my sleeping HR is 12% higher on average.” Specific, personal, and actionable.
The Practical Loop
- Tag consistently for 2–4 weeks.
- Open Patterns → find your highest-impact tags.
- Run a small experiment: eliminate the negative tag for two weeks and observe whether your Balanced/Elevated nights increase.
- Tag the change and let the data confirm or refute what you feel.
Root won’t tell you to stop drinking coffee. It will tell you whether coffee is actually affecting your sleep — with your own data.