The days before it tips
In bipolar disorder, the episode itself is not the real problem. The problem is that it usually takes you by surprise. A manic or hypomanic phase often feels good from the inside, especially at the start. More energy, less need for sleep, faster thinking. When you are in the middle of it, you rarely notice that something is tipping. The insight comes in hindsight, once the damage is already done.
This is exactly where data has its value. A phase shift rarely comes out of nowhere. It announces itself, often over days. The only problem is that these early signs are quiet, and memory smooths them over after the fact. Tracking makes them visible while they happen.
A systematic review by Jackson, Cavanagh & Scott (2003) in the Journal of Affective Disorders found that at least 80 percent of people with an affective disorder can name one or more prodromal symptoms, meaning early signs before the actual episode. Manic prodromes lasted longer than 20 days on average. That is a long window. Long enough to act, if you recognize it.
One important note up front: this text does not replace a diagnosis or treatment. Bipolar disorder belongs in specialist care, and tracking is an addition to that, not a substitute.
The two most reliable early warning signals
There are many possible early signs, but two come up again and again in research and in longitudinal data as the most robust.
1. Shortened sleep. Reduced need for sleep is one of the earliest and most reliable signs of a beginning (hypo)manic phase. Unlike normal tiredness, after little sleep you do not feel exhausted but awake and driven. Shortened sleep is not only a symptom here, but also a driver: sleep deprivation can fuel a phase further. A vicious circle.
2. Rising mood amplitude. Before a shift, the daily swings get larger. Mood fluctuates more sharply from day to day, sometimes within a single day. This growing range of fluctuation often shows up in the data before the average mood even breaks upward or downward.
What this looks like in the data
Growing amplitude before the phase shift
Look at the middle of the curve, not at the end. Long before mood shoots upward for good, the daily swings get wider. This growing amplitude is the real early signal. Anyone who only watches the average sees it too late.
What you should track
For a usable early warning system you do not need much, but you need it daily:
- Daily mood on a scale, ideally always at the same time
- Sleep duration (can be imported automatically via Apple Health)
- Energy and drive as a separate value, kept apart from mood
- Optional: Irritability, talkativeness, spending as context factors
The most important point: keep tracking during the stable phases too. A baseline from good weeks is what lets you recognize a deviation in the first place. If you only start at the first suspicion, you have no reference value.
(article: innerpulse/blog/2026/04/innerpulse-guide text: InnerPulse) reads sleep data automatically from Apple Health and places it next to your mood. That way you see shortened sleep and rising amplitude in the same timeline, without having to combine anything by hand.
Sleep is the lever, not just the indicator
One thing that sets bipolar disorder apart: here, sleep is not only a symptom you observe, but a factor you can turn. Because sleep deprivation can fuel a manic phase, recognizing shortened sleep early is doubly valuable. It is both a warning signal and a point of intervention.
If your data shows that you have slept noticeably less for three nights in a row and still feel wired, that is one of the clearest moments to check in with your clinician before the phase rolls in. How the sleep and mood relationship shifts in general is described in more detail in (article: innerpulse/blog/2026/01/how-sleep-affects-your-mood text: The sleep and mood lag).
An early warning plan you set in advance
Early signs are useless if you only start figuring out what to do once the emergency hits. What works is a simple, agreed-upon plan, often recorded with your clinician as a "crisis plan" or "early warning plan":
Example of an early warning plan
The data gives the traffic light colors a concrete trigger, instead of leaving them to a gut feeling that becomes unreliable anyway once a phase is beginning.
Important context
Three clarifications that matter here:
- Tracking does not replace diagnosis or treatment. Bipolar disorder belongs in specialist care. Your data is an addition, not a substitute.
- Data is not an early warning system at the push of a button. It only works with a baseline, and only if you set the plan in advance, while you are doing well.
- Not every swing is a phase shift. A bad day is a bad day. What matters is the pattern over several days, not the single value.
When it gets serious
Manic and depressive phases can both become dangerous, manic ones through risky behavior, depressive ones through suicidal thoughts. If you have suicidal thoughts during a depressive phase, get help. In the US, call or text 988 (Suicide and Crisis Lifeline). In the UK and Ireland, call Samaritans on 116 123. You can find a helpline for your country at findahelpline.com. If there is immediate danger to life, call your local emergency number (911 in the US, 112 across Europe) or go to the nearest emergency department. An acutely escalating mania with risky behavior (massive spending, dangerous risk-taking, complete loss of sleep) is also a reason to seek specialist or emergency medical help right away.
Data as an ally against your own perception
The treacherous thing about bipolar disorder is that self-perception becomes unreliable exactly when you need it most. A beginning hypomania does not feel like an illness, it feels like a good day. In that moment, tracking data is the one ally that does not lie along with you. It shows the pattern you cannot see from the inside.
Start while you are doing well. That is exactly when you build the baseline that warns you later.
Read on
- (article: innerpulse/blog/2026/02/recognizing-mood-patterns text: Recognizing patterns in your mood) shows how to read amplitude and trend cleanly.
- (article: innerpulse/blog/2026/01/how-sleep-affects-your-mood text: The sleep and mood lag) explains the connection between sleep and mood.
- (article: innerpulse/blog/2026/01/mood-journal-complete-guide text: Keeping a mood journal: The complete guide) is the tracking foundation.
- (article: innerpulse/blog/2026/03/is-my-medication-working-mood-data text: Is my medication working?) shows how to observe temporal connections between medication and mood in your data.
- Prodromes in affective disorders: Jackson, Cavanagh & Scott (2003)