Community-dwelling older adults, both male and female, aged 60 or above with MCI will be included. They will be recruited from community sites in Manila, Philippines through the assistance of local coordinators of the Office of Senior Citizens Affairs (OSCA) of Manila. The presence of MCI will be established following the criteria of Winblad et al. (2004) . A person is diagnosed to have MCI if 1) the cognitive function is not normal nor demented; 2) there is subjective report of the individual and/or informant on declining cognitive function compared to five years ago  on objective cognitive tasks; and 3) the performance of basic ADLs is preserved with only minor problems doing complex instrumental ADLs . Not normal cognitive function or presence of mild impairment will be based on Montreal Cognitive Assessment (MoCA) score of < 26 . Not demented will be based on not having a medical diagnosis of dementia or Alzheimer’s disease [5, 44]. Normal function in ADL and no or minimal impairment in instrumental ADL will be based on Katz ADL Scale  and Lawton Instrumental Activities of Daily Living (IADL) Scale . A trained neurologist-psychiatrist will examine the participants to provide the final diagnosis of MCI and give them the clearance to participate in the exercise. Only ambulatory older adults, with or without an assistive device, will be included since the interventions are community-based.
Older adults will be excluded if they have a diagnosis of dementia or Alzheimer’s disease; had history of major medical conditions such as cerebrovascular disease, cardiopulmonary condition, serious musculoskeletal disease, cancer, major psychiatric condition; have severe visual and/or hearing impairment; or are illiterate that will hinder them in participating fully and safely in the exercise and/or cognitive training programs. Furthermore, those who have been taking medications such as sedatives, antidepressants, diuretics, anti-epilepsy that might affect their cognitive function will not be included. Deliberately providing false information is a reason for termination of participation. Elevation of blood pressure beyond the expected increase due to exercise or the presence of fever and body pains for several days due to other causes may be reasons for withdrawal to participate. To enhance adherence to the intervention, participants will be receiving a travel allowance and light refreshments every time they attend sessions.
Study design and procedures
This proposed study is a randomized controlled trial using a 4-group design including three intervention groups (physical training alone or cognitive training alone, or combined physical and cognitive training) and one waitlist control group. The study will comply with the principles of the Declaration of Helsinki 2013 and Good Clinical Practice Guidelines. Ethics approval was secured from the Human Subjects Ethics Sub-committee of the Hong Kong Polytechnic University-Department of Rehabilitation Sciences (HSEARS20170402001) and the Ethics Review Committee of the University of Santo Tomas-College of Rehabilitation Sciences (FI-2017-002). The SPIRIT guidelines were followed in the design of this study protocol.
Trained study personnel will interview the participants using questionnaires to obtain their demographic information and medical history, physical activity level and history of falls, and will conduct cognitive and physical performance tests to determine their eligibility to participate in this study. Only those who signed the informed consent will be included in this study.
Random allocation will be carried out by a third party not involved in the study by random draw to blindly allocate each participant to one of the four groups. To minimize experimental contamination through social interaction and communication among the participants, the interventions will be delivered at different sites.
Measurement points include baseline/pre-intervention, post-intervention, and six months follow-up assessments and will be conducted by trained study personnel blinded to the randomization assignment. Physiotherapists and occupational therapists who will be administering the interventions will not be involved in the outcome assessment. They will also have workshops and discussions to standardize the administration of the different training programs. After the intervention period, aside from a calendar-diary, monthly phone calls will also be done to follow up the participants about their fall status, physical activity, and exercise behavior. Figure 1 illustrates the study flow. Appendix shows the timeline of the study.
Sample size computation
The sample size and power calculations will be based on fall rate which is the primary outcome of the study. Using statistical software Gpower 188.8.131.52, a priori power analysis was conducted using the computed effect size (Cohen’s d = .39) from Trombetti et al. (2011) . Assuming 80% power with 5% Type 1 error and four groups, the calculated total sample size is 80. Considering a 16% dropout rate (Trombetti et al., 2011) , the total number of participants is inflated to 93. There should be at least 23 participants per group.
Physical training alone (PT)
The participants in the PT group will be performing a series of physical exercises as a group supervised by trained physiotherapists. To ensure proper guidance and safety in the performance of the exercises, there will be one physiotherapist for every five participants.
The exercise program will start with 5–10 min of warm-up including calisthenics and general flexibility exercises, followed by 60–90 min of multicomponent exercise programme including endurance, strength and balance exercises (with a focus on balance training), and will end with 5–10 min of cool down including calisthenics and general flexibility exercises. This will be done three times per week over 12 weeks to achieve the recommended minimum accumulated 40 h of exercise . The exercises will be progressed individually, increasing the repetitions or sets first before the resistance, to maintain a moderate level of exercise intensity which is 5 to 6 on a scale of 0 to10 for the level of physical exertion . Rest periods will be provided whenever necessary. Table 1 shows the types of exercises included in this program.
The exercises were based on the recommendations of Sherrington and Tiedemann (2015)  and were selected because of the minimal use of sophisticated equipment which is important in an exercise regimen designed for a community setting. The participants will be given brochures containing simplified exercise instructions with illustrations and printed using a large font size to help them remember the exercise routines and encourage them to perform the exercises at home.
Cognitive training alone (CT)
The participants in the CT group will be involved in a set of paper-based cognitive exercises as a group supervised by trained occupational therapists. To provide close monitoring and immediate feedback during the training, there will be one occupational therapist for every five participants.
The intervention will start with 5–10 min of warm up to give the instructions, followed by 60–90 min of cognitive training designed to train specific cognitive functions such as orientation, memory, attention and executive function, with emphasis on executive function which is associated with increased risk for falls in older adults , and will end with 5–10 min of cool down for feedback and processing of responses. This will be done once a week for 12 weeks which is the optimal duration of cognitive training . Even though the cognitive training is delivered in a group, the difficulty level and progression will be individualized. Rest periods will be provided whenever necessary.
Table 2 shows the contents of the cognitive training. The contents of training for orientation and attention were based from Brum et al. (2009) , and Vojtkofsky and Feldman (2015) . For memory, training will include rehearsal, association, visual imagery and concentration. The use of multiple strategies in memory training is chosen as larger improvement is expected with this approach compared to single strategy approaches [50, 51]. For executive function, training will consist of extensive repeated practice of practical tasks to simulate usual daily activities instead of relying on strategies to strengthen cognitive processes. This is founded on the process-based approach which has been proven to be highly effective in improving executive function .
The paper-and-pencil format of cognitive training is selected in consideration of the low economic situation of the community. The participants will be given educational brochures with large fonts and illustrations to inform them about the importance of maintaining or improving their cognitive abilities citing specific ways to be mentally active in old age.
Physical and cognitive training (PACT)
The participants in the PACT group will be performing activities that integrate cognitive training in physical exercise routines as a group supervised by trained personnel either physiotherapists or occupational therapists. To ensure proper guidance, safety, close monitoring and immediate feedback about their performance of the activities, every five participants will be supervised by one therapist.
The PACT programme will start with 5–10 min of warm-up including calisthenics and general flexibility exercises, followed by 60–90 min of exercises similar in design, intensity and progression with the PT group but with cognitive components incorporated in each type of exercise, and will end with 5–10 min of cool down including calisthenics and general flexibility exercises and a post-session discussion to recall the activities performed in the just-concluded session. This will also be done three times per week over 12 weeks. Rest periods will be provided whenever necessary. Table 3 presents the design of the PACT program. The participants will likewise be given the same brochures given to the PT and CT groups for additional information and reference.
Waitlist group (WG)
The participants in the WG will serve as the control group on waitlist. They will be instructed to go on with their usual daily routine and will receive the intervention, combined physical and cognitive training, at a later date.
The following primary and secondary outcome measures will be performed by trained assessors blinded to the group allocation. Measurements will be taken at baseline, at the conclusion of the intervention period, and 6 months post-intervention. Additional information regarding the exercise behavior and the physical and social activities participated in by the eligible participants in the last 6 months will be obtained during the 6 months post-intervention follow-up assessments .
Primary outcome measure
Each participant will be given a calendar diary to mark any incidence of fall on a weekly basis during the assessment period. A relative or caretaker of the older person will be asked to validate the information. The confirmation by another person is done because self-reporting techniques, particularly for individuals with cognitive impairment, may not be accurate due to recall bias . Monthly phone calls will be done to update the falls history of each participant.
Secondary outcome measures
The secondary measures will be assessments that evaluate overall fall risk, cognitive, physical and psychological factors related to falls.
Overall fall risk
The overall fall risk will be determined using the short-form Physiological Profile Assessment (PPA) and the Falls Risks for Older People in the Community (FROP-Com) Screen.
The PPA short form is a valid, reliable and objective series of tests used to evaluate the physiologic risks of falls and to classify fallers from non-fallers . It is composed of five subtests on 1) visual contrast sensitivity, 2) proprioception, 3) hand reaction time, 4) knee extensor muscle strength, and 5) postural sway . Data from these tests will be encoded in a web-based software program such that the performance of one participant is compared to a normative database to determine whether the individual has a low or high risk of falls .
The FROP-Com Screen is a brief screening tool used to determine older persons who are prone to fall . It is composed of three items on 1) the number of incurred falls within the past 12 months, 2) the level of dependence in doing instrumental ADLs, and 3) balance in walking and turning of older adults . A score of 0–3 means the older adult has a low risk for falls, while a score from 4 to 9 denotes a high risk for falls .
Montreal Cognitive Assessment (MoCA) is a brief 10-min, 30-item, one-page cognitive screening, diagnostic and tracking tool with high sensitivity and specificity for detecting MCI [43, 57]. It is widely used internationally and has been translated into many languages. It assesses several cognitive domains such as memory, executive function, attention, language, abstraction, naming, delayed recalls, and orientation.
The Memory Index Score of the Montreal Cognitive Assessment (MoCA-MIS) consists of 15 items on memory and validated to help determine in predicting conversion to Alzheimer’s disease or dementia from MCI over 18 months . It is calculated by getting the sum of the number of words correctly remembered by the participant in free delayed recall, category cued recall, and multiple choice-cued recall multiplied by 3, 2 and 1, respectively, to obtain a score which ranges from 0 to 15 .
Executive Function Performance Test (EFPT) is a valid and reliable instrument that uses structured cueing and scoring system to examine executive function in doing basic real-world tasks like hand-washing, oatmeal preparation, telephone use, taking medication, and paying bills . To consider the local context, only the use of the telephone and taking of medication will be performed in this study.
The Timed Up and Go Test (TUGT) will be used to examine the dynamic balance of the participants. It is a simple and reliable  test that measures the time in seconds it takes a person to stand up from a chair, walk a distance of three meters, turn around, walk back towards the chair and sit down again. It is a sensitive and specific tool for identifying older adults who are at risk of falls . Those who take longer than 14 s to complete this test have a high risk of falls .
Gait speed will be measured using the 10-Meter Walk Test (10MWT). This is a valid and reliable gait speed assessment in older adults . It is the quotient of distance covered in meters, and the duration in seconds it takes the participant to walk that distance. The participants will perform the test at their preferred speed and fastest speed possible [62, 63].
The strength of the lower limb muscles will be assessed using the 30s-Chair-Stand Test (CST). This is done by asking the participants to perform sit-to-stand on a standardized chair as many as they can within 30 s. The CST has excellent validity and reliability in community-dwelling older persons .
Fear of falling
The Falls Efficacy Scale – International (FES-I) will be used to assess the fear of falling of the participants. The tool is composed of 16 items. Each item is scored based on a four-point scale (1 = not at all concerned; 2 = somewhat concerned; 3 moderately concerned; and 4 = very concerned). The score ranges from 16 to 64. A high score means more fear of falling. FES-I has been shown to have excellent psychometric properties including construct validity, internal consistency, and test-retest reliability and discriminatory power when used in older persons [65,66,67,68].
Quality of life
Perceived Well-Being (PWB) Scale is a brief and easy to use tool that measures the quality of life specifically in the domains of psychological and physical well-being. It has high internal consistency and validity and is significantly correlated with variables that affect an individual’s well-being . The score ranges from 16 to 116. A high score means greater perceived well-being.
EuroQoL-5 dimensions-5 levels (EQ-5D-5 L) is a valid and standardized tool developed to provide a simple, generic measure of health status for clinical and economic analysis . The five dimensions are mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension has five levels: no problems, slight problems, moderate problems, severe problems, and extreme problems . The tool has shown evidence of responsiveness to change . It has been translated into over 120 languages and the validated Tagalog version of EQ-5D-5 L will be used with permission in this study.
Descriptive statistics will be used to summarize demographic data using mean standard deviations, and percentages. Baseline values across the groups will be compared using one-way ANOVA for interval/ratio data and Kruskal-Wallis test for nominal/ordinal data. The multivariate repeated measure ANOVA will be used to determine the pre- and post-intervention effects and differences among group data with a normal distribution; otherwise the Friedman test will be utilized. The incidence rate ratio, computed as the number of falls divided by the duration of falls monitoring for every participant, will be used to compare intergroup fall rate. An intention-to-treat analysis will be used for missing data due to dropouts. The p-value of < 0.05 is considered significant. Data analysis will be performed using SPSS version 23 for Windows.
Collected data using a paper-based assessment kit will be encoded in MS Excel by one study personnel blinded to the group allocations. This will be double-checked by another research assistant. The privacy and confidentiality of the participants will be kept by making no personal information available to the public. Signed consent forms and filled data collection sheets will be stored in a locked cabinet. The encoded data will be kept in a password-protected computer folder. Data will be stored for five years and will then be deleted. In the event of any publication involving this study, the participants’ identities will remain confidential. Aside from the researchers, the study monitors from the ethics committee will be granted monthly direct access to the records for purposes of procedure and data verification and interim analyses. Any modifications in the study procedure will be reported to the Human Subjects Ethics Sub-committee of The Hong Kong Polytechnic University-Department of Rehabilitation Sciences and the Ethics Review Committee of the University of Santo Tomas-College of Rehabilitation Sciences, and to ClinicalTrials.gov. To disseminate the findings of this research, the results and implications will be reported to the participants and presented formally in professional conferences, and published in a reputable journal.
Participants might experience fatigue while doing exercise, and therefore adequate rest periods and breaks will be provided as needed. Vital signs will be constantly monitored before, during and after the interventions every session. Light refreshments will be provided to prevent dehydration. In some activities that will challenge their balance, trained therapists will be close by to provide support to prevent them from falling. The calendar-diary will also be used to monitor changes in health status and record adverse events every session during the intervention period and then every month after the intervention phase. In the case of any untoward incident like dizziness, injuries, or falls during the conduct of the study, appropriate first aid, medical attention or referral to a clinic or hospital, will be immediately done to the participant as soon as reported.