Au 2019.

Study characteristics
Methods	Design: Parallel randomised controlled trial. No. of participating centres: Two: United Christian Hospital and Prince of Wales Hospital, Hong Kong, China. Study dates: January 2015 to June 2016.
Participants	Inclusion criteria: Primary caregivers (spouse, kin, or sibling) aged 25 or above caring for a care recipient diagnosed with mild to moderate Alzheimer's disease. Providing at least 14 hours of care per week for at least 3 months. Exclusion criteria: Presence of signs of the following conditions: severe intellectual deficits, suicidal ideation, psychotic disorders, or lack of the ability to read Chinese and speak Cantonese. No. of participants randomised to intervention: ‐ TBA (n=56). ‐ TGM (n=55). Baseline characteristics: ‐ Caregivers: Mean age was 57.1 years (SD=10.2); 90 females and 21 males. The average years of caregiving were 4 (1.8) with an average of 10.9 (7.6) hours/day spent in caregiving. Mean score of burden (22‐item ZBI) at baseline 30.9 (SD=17.2). ‐ Care recipients: 69 participants (62%) presented moderate dementia, 42 participants (38%) presented severe dementia.
Interventions	Experimental: Telephone‐administered psycho‐education with behavioral activation intervention (TBA). Control: Telephone general monitoring. Length of intervention: Four‐month.
Outcomes	Primary outcomes: ‐ Depressive symptoms (CES‐D). Secondary outcomes: ‐ Caregiver burden (Zarit Burden Scale). Outcomes assessment: An average of about 5 months since baseline.
Notes	Clinical trial registration code: NCT03552159. Funding: Research Council of the Food and Health Bureau of the Government of the Hong Kong Special Administrative Region (Health Care Promotion Fund 0310015). Conflicts of interest: None declared. Fidelity / adherence: Authors report that 60% of participants fully covered the program and 40% adequately covered the program.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Allocation sequence was obtained by random number generation by a staff who was not involved in enrolling/assigning participants... block randomization was used to achieve balance between the numbers of participants in both arms".
Allocation concealment (selection bias)	Low risk	Quote: "The allocation was concealed in sequentially numbered sealed envelopes".
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "... both assessors and participants were blinded to intervention condition".
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Assessments were carried out by research assistants who were blind to the group allocation of participants". Comment: Intervention components and procedures were comparable.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: The study presents reasons for losses to follow‐up without relevant differences between arms, also ITT analysis was done.
Selective reporting (reporting bias)	Low risk	Comment: Study registration available (NCT03552159), all prespecified results are reported.
Other bias	Low risk	Comment: No other source of bias detected.