Caregiver adverse events in Tele-STELLA were often physical, but the psychological burden stood out

Family care partners caring for people living with Alzheimer’s disease and related dementias carry a health burden that often remains invisible until a crisis surfaces. In the Tele-STELLA study, adverse event narratives offered a window into that burden by capturing not only medical problems, but also emotional strain, conflict, and distress that can emerge while caring for a cognitively impaired relative. Investigators reviewed 43 adverse events reported by 36 care partners and sorted them into Physical, Psychological, or Random/Other categories. The distribution was not even: Physical events accounted for 48.8% of adverse events, Psychological events for 39.5%, and Random/Other events for 11.6%.

Infections dominated the physical category. Stress and distress made up half of the psychological events.

That pattern matters because caregiver safety reviews usually focus on obvious medical harms, while the emotional consequences of caregiving may pass as background noise. Here, the narrative structure of the adverse events itself turned into clinical data. The presence of infections points to the wear-and-tear of prolonged caregiving, reduced sleep, disrupted routines, and the practical strain of managing someone else’s complex needs while neglecting one’s own health. Psychological events, by contrast, suggest that emotional load is not a side issue; it can be one of the most frequent adverse outcomes in a caregiving intervention study.

Baseline grief and reactivity tracked with psychological adverse events

The investigators also compared baseline psychological characteristics between care partners with and without each adverse event category and subcategory. Two measures were especially informative: baseline depressive symptoms and grief, along with behavioral reactivity. Care partners who experienced psychological adverse events had higher baseline grief and behavioral reactivity than those who did not. When the psychological events were broken down further, the signal became sharper. Stress- and distress-related adverse events were associated with higher depressive symptoms and higher reactivity, while aggression-related adverse events were associated with greater grief.

These associations are clinically plausible. Grief in dementia caregiving does not wait for death; it often begins with the slow loss of the person’s memory, autonomy, personality, and shared routines. A care partner who enters an intervention with elevated grief and reactivity may be more vulnerable to episodes of emotional overload, especially if the caregiving relationship is already strained by behavioral symptoms, communication breakdowns, or escalating dependence. Aggression-related events likely reflect the acute stress of managing difficult interactions, but they may also expose a deeper emotional context in which the caregiver feels isolated, threatened, or unable to protect the person they care for. Depressive symptoms and behavioral reactivity, meanwhile, can amplify each other. A caregiver who is already emotionally depleted may interpret routine demands as more aversive and respond with greater physiological and psychological strain.

A descriptive signal, not a causal claim

This was a descriptive, exploratory analysis, and the findings do not establish causality or predictive validity. The study design cannot show whether baseline grief and reactivity led to the adverse events, or whether the events reflected a preexisting vulnerability that the baseline measures only partially captured. Small numbers also limit how far the subgroup findings can be pushed. Forty-three adverse events is enough to sketch a pattern, but not enough to build a definitive risk model or to sort out which emotional factors matter most across the caregiving trajectory.

Even so, the structure of the data has practical implications for how caregiver interventions are monitored. Conventional assessment tools tend to emphasize depression, burden, or quality of life at scheduled visits. Narrative adverse event capture adds another layer, especially in studies like Tele-STELLA, where the lived experience of the caregiver may shift between emotional distress, acute conflict, and ordinary physical illness. A caregiving intervention can look acceptable on paper while a participant is quietly struggling with sleep disruption, infection, or rising distress that never fits neatly into a checkbox.

Why narrative adverse events could improve caregiver monitoring

Tele-STELLA’s approach suggests that adverse event classification may complement standard caregiver health assessments. In a dementia care context, the caregiver is often treated as the support structure rather than as a patient with clinical needs of their own. That framing misses the point that caregiving itself can produce measurable harm. By organizing event narratives into physical and psychological domains, the study created a more complete picture of risk during participation in a telehealth-based dementia support program.

The practical value lies in early detection. If stress/distress-related events cluster among care partners with higher depressive symptoms and reactivity, then baseline screening could help identify who needs more intensive support, tighter follow-up, or additional emotional scaffolding during an intervention. If aggression-related events align with higher grief, then caregiver support may need to address anticipatory loss and relational strain, not only symptom management in the person with dementia. In that sense, the narrative data do more than describe adverse events; they point to a way of reading caregiver vulnerability before it becomes a dropout, a crisis, or an unreported health decline.

The study was approved by the Oregon Health & Science University Institutional Review Board, and the trial registration is listed at NCT04627662. The preprint remains unreviewed, so the findings should be treated as hypothesis-generating rather than practice-changing. Even so, the core message is hard to ignore: in dementia caregiving research, adverse events are not limited to physical illness, and the emotional terrain may be where the earliest warning signs appear.


Reporting by Dr. A. Nikhil Krishnan, MBBS for Medicine World Foundation, with additional data provided by medRxiv (Clinical Preprints).
PRIMARY CLINICAL EVIDENCE & REGISTRY DATA