The PICMOA-2 trial is an open-label, 6-month RCT with 2 parallel groups, including 1 intervention group and 1 passive control group. The overall timeline of the procedure is shown as a SPIRIT flow in Table 1. The main and secondary outcomes will be assessed at baseline and after a 6-month intervention period.

We recruited community-dwelling older adults aged 65 years and older without cognitive decline living in Kishiwada city, Osaka Prefecture, Japan, near a metropolitan area with a population of 188,129 as of December 2023 [28]. This study’s design is illustrated in Figure 1. This procedure included eligibility screening, informed consent, outcome assessments, randomization, and 6 months of web-based conversational intervention. The inclusion criteria are as follows: a Mini-Mental State Examination (MMSE) score ≥ 24, available to provide written consent, and able to undergo the required tests and interventions on specified dates. We screened medical conditions and medication use and excluded patients with dementia; those with a history of disease or previous medications that could affect the central nervous system, such as neurological impairment, serious complicating disorders, and a history of serious head injury; and those certified as having a long-term care condition (“care needs levels” or “support need levels”) by the national public long-term care insurance system in Japan.

Participants were recruited by sending recruitment letters and flyers to community-dwelling older adults from the municipal government and posting announcements for enrollment in a new community paper. We mailed 3000 recruitment letters to randomly selected older adults without certificates of long-term care in 4 living areas of Kishiwada city. Some flyers were also distributed to the public and in collaboration with organizations. Once potential participants expressed interest in participating in the trial, they were screened for eligibility. Participants in both groups will receive cash compensation.

The intervention program is PICMOA-2, a web-based conversational intervention that uses a PC tablet at the participant’s home. Participants are instructed to take photos in their daily lives as much as possible according to the specific 12 themes during the 6-month intervention period and join conversational intervention based on PICMOA-2 once every 2 weeks for 6 months. Figure 2 describes the intervention structure of PICMOA-2. Participants are required to upload to the web system using a smartphone app that we originally developed before the intervention session, join the group conversational intervention session, and have a presentation and discussion about the photos that participants took during the session.

In the conversational session, each participant in a designated group of 4 gives a 1-minute speech about the 2 photos they took in their daily lives. Other participants then ask questions related to the photos, and the presenter answers the questions in the context of a natural conversation for 2 minutes for each photo. Research staff administer intervention using the PICMOA-2 system and support participants’ PC operations through remote access.

In the case of dropouts, substitutes are included in the group conversation to maintain the intensity of the intervention since different numbers of group members can lead to quantitative differences in the amount and duration of speaking, potentially impacting the effect sizes.

Smartphones for taking photos and PC tablets (Surface Go; Microsoft Corp) for conversational sessions with an internet connection were prepared for all participants.

We used G*power [38] to calculate the sample size, assuming the following conditions: 2-sided hypothesis test; medium effect size (f=0.25); 95% power; an analysis of variance model between groups over time; and 5% α level. Therefore, 65 participants are required, with a 20% dropout rate.

Stratified block randomization, a block design (2, 4, and 6 block sizes) with a 1:1 allocation, was performed using the University Hospital Medical Information Network Internet Data and Information System for Clinical and Epidemiological Research Cloud version [39]. The participants were allocated to either the intervention or the control group. Sex (male or female) and MMSE scores (28 or higher and 27 or lower) at baseline were used for stratification. KWM is responsible for the randomization. As this study uses an open-label design, we aim to prevent potential assessment bias by blinding the assessors.

This study evaluates the effects of the PICMOA-2 intervention on outcome measurements. Linear mixed models with random intercepts for outcome measurements will be used to estimate the intervention effects. Our model will include the group assignment factor, time factor, and group × time interaction term and evaluate the intervention effect from the interaction term. As a secondary aim, we will test whether the intervention effects varied with the covariates at baseline to examine the effective target attributes.

All analyses will be conducted using the R software (The R Foundation). Statistical significance will be set at P<.05.

This trial was approved by the Ethics Committee of RIKEN (RIKEN-W1-2022-063) and was registered in the University Hospital Medical Information Network clinical trial registry (UMIN000050877).

This protocol aimed to evaluate the effect of a 6-month PICMOA-2 intervention, a web-based group conversational intervention, on cognitive function, frailty, and social and psychological indicators to determine the effectiveness of a web-based remote conversational intervention that we originally developed.

This study has several strengths. First, it may have clinical significance by accumulating evidence on the effectiveness of social interventions on cognitive function and digital health practices. Although large-scale observational studies have demonstrated a link between social interaction and cognitive health [2-7], whether social interventions can slow cognitive decline remains unknown. Since there is low evidence of social intervention and there are no clear clinical guidelines, this study may add knowledge on the effectiveness of social intervention. Second, this study examined whether there are differences in intervention effects depending on attributes by collecting subjective and laboratory data. A previous study in a completed clinical trial on mild Alzheimer disease suggested that baseline plasma NfL, which is a neurological biomarker, holds independent information on short-term cognitive decline [40]. Our previous RCT of a face-to-face conversational intervention program also found a positive intervention effect in the group with lower levels of NfL [23]. In addition, our previous PICMOA trial using smartphones suggested that familiarity with digital devices causes differences in the intervention effect [27]. This research may contribute to the search for clues to effective target images for remote social intervention by examining effective targets. Finally, this study directly recruited community-dwelling older adults mainly via mail in cooperation with the municipal government. This method can eliminate selection bias and enhance generalizability to real community intervention settings compared with recruiting from specific organizations or surveys.

This study uses an open-label trial design. Although the assessors are blinded to avoid assessment bias, blinding the research staff and participants is not possible for such behavioral interventions. In addition, the sample size was designed with a minimum of 65 participants for feasibility considerations. Large-scale validation is required to examine the effectiveness of the intervention.

This protocol outlines the design of an RCT study that evaluates the effect of a 6-month intervention with PICMOA-2 on cognitive function, frailty, and social and psychological indicators. Though examining the effect of the PICMOA-2 program, this study will provide evidence of the effectiveness of social interventions on cognitive function and identify effective target images for remote social intervention.