Щоденник виявлення тригерів нападу — logging sleep, medication, stress and alcohol day by day to statistically uncover each patient's personal seizure trigger signature
Roughly nine in ten people with epilepsy can name at least one factor they believe precedes their seizures. Decades of self-report surveys and controlled sleep-deprivation studies converge on a short list of recurring categories: insufficient sleep, missed anti-seizure medication, alcohol (especially withdrawal), psychological stress, hormonal fluctuation across the menstrual cycle, flashing or high-contrast visual stimuli, and systemic illness or fever. A trigger diary begins by teaching the patient to watch for exactly these seven signals.
Sleep deprivation: Even a single night of shortened or fragmented sleep measurably lowers cortical excitability thresholds. Sleep-deprived EEG remains a standard clinical activation procedure precisely because sleep loss reliably provokes epileptiform discharges in susceptible patients.
Missed medication doses: Sub-therapeutic anti-seizure drug levels are one of the single largest preventable causes of breakthrough seizures. Even one skipped dose can drop plasma levels below the threshold needed for seizure control, particularly for short-half-life drugs.
Alcohol: Acute intoxication and, more strongly, alcohol withdrawal both lower seizure threshold via GABAergic and glutamatergic rebound effects. "Weekend seizures" after heavy drinking are a recognizable clinical pattern.
Stress: Psychological stress is the most frequently self-reported trigger in survey studies, though its physiological pathway (cortisol, autonomic arousal, sleep disruption) is harder to isolate than the others.
Catamenial pattern: Cyclical estrogen/progesterone shifts modulate neuronal excitability; many women with epilepsy notice seizure clustering around menstruation or ovulation.
Photosensitivity: Flashing lights, high-contrast patterns, and certain screen content provoke generalized discharges in a genetically predisposed subset of patients, most commonly with juvenile myoclonic epilepsy.
Illness / fever: Elevated body temperature and systemic inflammation lower seizure threshold in both children and adults, independent of the classic pediatric febrile-seizure syndrome.
Self-reported triggers are a useful starting hypothesis, not a diagnosis. Prospective studies that ask patients to log a suspected trigger and their subsequent seizures, then test the association statistically, routinely find that only a minority of self-nominated triggers hold up under scrutiny — patients frequently confuse a trigger with a prodromal symptom (the trigger being felt because a seizure was already building, rather than causing it). This is precisely the gap a structured diary and correlation analysis are designed to close: turning a list of plausible categories into a personally verified profile.
A landmark prospective diary study (Haut et al., Neurology, 2007) tracked self-predicted seizure days against actual EEG-confirmed seizures and found that most patients could not reliably predict individual seizure days from mood or stress alone — reinforcing why systematic, dated logging outperforms memory and intuition.
A trigger diary only works if it is filled in consistently. Each day the patient answers the same short set of questions — how many hours did I sleep, did I take every dose, how alcohol did I drink, how stressed do I feel on a 1–10 scale — and logs whether a seizure occurred, at what time, and what type. Over weeks this produces a day-by-day matrix that can be mined for statistical structure rather than relying on memory at the next clinic visit.
1. Sleep — hours slept the previous night, plus any awakenings 2. Medication — every scheduled dose taken (yes/no per dose, not just per day) 3. Stress — a simple 1–10 self-rating, logged at a fixed time to reduce recall bias 4. Alcohol — standard drink units consumed 5. Seizure events — occurred yes/no, time of day, seizure type/semiology, and duration if known
Optional fields commonly added later include menstrual cycle day, screen/flashing-light exposure, illness or fever, and sleep quality (not just quantity). Keeping the core five short is deliberate: diaries with too many fields see completion rates collapse within a few weeks.
Electronic patient-reported outcome (ePRO) apps with a daily push reminder consistently out-perform paper diaries on completion rate, and they timestamp entries so retrospective backfilling (a major source of recall bias) is discouraged or blocked entirely.
Even with perfect completion, diaries systematically under-count seizures. Studies comparing self-reported diaries against continuous video-EEG monitoring in epilepsy monitoring units find that patients miss a substantial fraction of their own seizures — nocturnal events, subtle focal-aware seizures without dramatic motor signs, and events during sleep or inattention are the most commonly missed. This is an important caveat for every later stage of the pipeline: the diary is a proxy for seizure occurrence, not a perfect ground truth.
Video-EEG validation studies have repeatedly shown that patients under-report complex partial and nocturnal seizures by roughly 50%, which is why modern trigger-diary programs increasingly pair self-report with wearable motion/EEG sensors to catch what the diary alone misses.
Once several months of daily entries exist, the diary becomes a dataset. Each logged factor is tested against seizure-occurrence days using lagged windows — did sleep the night before matter more than sleep two nights before? — and statistical association measures separate a real personal trigger from coincidence. This step converts a list of "possible" triggers into a small set of factors that actually track this patient's seizures.
Self-controlled case series (SCCS): compares the rate of seizures during "exposed" periods (e.g., nights following <5 hours sleep) against the same patient's "unexposed" periods, controlling for individual baseline risk without needing a separate control group.
Lagged logistic / Poisson regression: models the odds or rate of a seizure on day t as a function of sleep, adherence, stress, and alcohol on days t-1, t-2, t-3, letting the analysis find which lag each factor operates on.
Generalized estimating equations (GEE): account for the fact that daily observations from the same patient are correlated with each other, avoiding artificially inflated significance.
Simple point-biserial or Pearson correlation on rolling weekly aggregates is often used as a first coarse pass before the more rigorous lagged models are applied.
Seizures are relatively rare events even in poorly controlled epilepsy, so the number of "exposed" days that actually precede a seizure is small — statistical power is often the binding constraint, not the analysis method. Testing seven trigger categories across several lag windows multiplies the chance of a false-positive correlation, so correction for multiple comparisons is essential. Confounding is also pervasive: poor sleep often co-occurs with high stress and, in some patients, with reduced medication adherence, making it hard to isolate a single causal factor without a large enough diary.
Privitera and colleagues' prospective study of self-reported seizure precipitants found that stress — the single most commonly nominated trigger in surveys — showed only weak and inconsistent statistical association with actual seizure timing when tested rigorously, underscoring why correlation analysis, not intuition, must decide which triggers are real for a given patient.
Correlation analysis rarely returns a single trigger — it returns a weighted profile. One patient's seizures track almost exclusively with sleep loss; another's track with stress and alcohol combined; a third shows no single dominant factor but a summation effect, where several mild stressors together cross a personal seizure threshold that none would cross alone. Converting the statistics into this individual profile is what makes the diary clinically actionable.
Most modern trigger models treat seizure likelihood as a threshold problem rather than a single-cause one: each risk factor (poor sleep, low drug level, high stress, alcohol) shifts an underlying seizure-susceptibility curve upward by a small amount. A seizure becomes likely once the combined shift crosses the patient's personal threshold for that day. This explains a common clinical observation — mild sleep loss alone rarely triggers a seizure, but the same mild sleep loss on a day with also-missed medication and high stress often does. The diary's weighted profile is essentially an estimate of how much each factor contributes to that day's combined shift for this specific patient.
Patient A: correlation analysis shows a strong, consistent association between nights under 5 hours of sleep and next-day seizures, with adherence and stress showing negligible independent effect once sleep is accounted for. Her action plan centers almost entirely on sleep hygiene and a fixed sleep schedule.
Patient B: no single factor dominates, but the combination of stress ≥7/10 and alcohol on the same day shows a strong association, while sleep alone shows none. His plan targets stress management and alcohol moderation together, not sleep.
The same seven-category checklist from Stage 1 produces two completely different, individually actionable profiles — which is exactly why generic "avoid stress and sleep well" advice under-performs a diary-derived, personalized plan.
Because trigger weighting is patient-specific, clinical guidelines increasingly recommend that every newly diagnosed patient complete a structured 3-month diary before trigger-avoidance counseling is finalized, rather than issuing the same generic advice to everyone.
Identifying a trigger profile only has value if it changes behavior and that behavior change measurably reduces seizures. The final stage of the diary loop is intervention — sleep hygiene, adherence support tools, stress-reduction training, alcohol moderation — matched to each patient's confirmed profile, with the same diary continuing afterward to verify whether seizure frequency actually falls.
Sleep-dominant profile: fixed sleep/wake schedule, screen curfew, treatment of comorbid insomnia or sleep apnea, and — when relevant — shift-work adjustment.
Adherence-dominant profile: pillbox reminders, blister-pack organization, simplified once- or twice-daily dosing regimens, and automated refill alerts.
Stress-dominant profile: cognitive behavioral therapy, mindfulness-based stress reduction, and, where feasible, workload or schedule adjustments during identified high-risk periods.
Alcohol-associated profile: structured moderation goals and, where appropriate, referral for alcohol-use support, since withdrawal — not just intoxication — is often the more potent seizure risk.
Catamenial pattern: perimenstrual dose adjustment protocols developed with the treating neurologist.
Photosensitivity: screen-filter settings, avoidance of known triggering content, and polarized lenses where indicated.
The diary does not stop once an avoidance plan starts; it becomes the outcome measure. Monthly seizure frequency before the intervention is compared against the same measure over the following months, using the identical logging format so the comparison is apples-to-apples. Self-management education programs that combine trigger identification with structured avoidance planning have reported seizure-frequency reductions in the range of 30–40% over 6–12 months of follow-up, though results vary widely with baseline severity and how consistently the plan is followed. A sustained downward trend across several months of continued diary entries is the strongest available evidence, short of formal clinical trial data, that the identified triggers were real and the avoidance strategy is working.
Structured epilepsy self-management programs (such as the MOSES education curriculum) that pair trigger-diary training with targeted behavioral coaching have reported seizure-frequency reductions of roughly one-third in participating patients, alongside improved medication adherence and quality-of-life scores.