Description

Study design A hybrid study comprising of a cross sequential study and a randomized controlled trial.

Study area The study locations are the following five peri urban communities out of the ten peri urban communities in the La Nkwantanang Madina Municipality (LANMA) of the Greater Accra Region in Ghana; Ayi Mensah, Kweiman, Danfa, Adoteiman and Otinibi. These have been conveniently selected because they are a continuous stretch of five communities in the order listed above to reduce the cost of transportation of data collectors and logistics, distribution of the intervention and data collection based on funds available for the project.

Sample size Assumptions: A Cohen's medium effect size of 0.3, a statistical power of 80% at a 5% level of significance, G-power 3.9.1.7 software generated a sample size of 90 participants per treatment group. Accounting for attrition rate + loss to follow up of 10% = 99 participants \* 2 treatment groups

• A total sample size of 198 elderly individuals. This has been rounded off to 200 participants. The census to recruit the 200 eligible participants has been completed as of December 13, 2024.

A cross-sectional survey to identify the relationship between usual dietary intake, handgrip strength and blood pressure has been conducted for the 200 participants. Eighty of the participants will be randomly selected as participants for a 12 week randomized control trial.

Assumptions (80 participants): A Cohen's medium effect size of 0.25, a statistical power of 90% at a 5% level of significance, GPower 3.1.9.4 generated a sample size of 36 participants per treatment group. Accounting for attrition rate and a loss to follow up of 10% = 40 participants \* 2 treatment groups. A total sample size of 80 elderly individuals.

Data collection for the randomized control trial An interviewer administered questionnaire will be used to obtain the following information at different points of assessments

Baseline data collection Sociodemographic characteristics and medical history such as age, sex, highest level of education, marital status, ethnicity and use of iron supplements, anti-hypertensive drugs and cholesterol lowering drugs will be collected.

Baseline, midline (6 weeks) and endline (12 weeks) A multiple pass quantitative 24-hour dietary recall to obtain information on dietary intake. Household measures and food models will be used to estimate portion sizes. The Physical Activity Scale for the Elderly (PASE) by Washburn et al. (1993) will be used to measure level of physical activity. Body weight, hand grip strength, systolic and diastolic blood pressure will be measured.

Weekly All participants will receive nutrition education at baseline and weekly thereafter. Information on adherence will be collected by evidence of pouches of previous week's ration to calculate percentage of quantity consumed. The 24-hour dietary recall will also be used to track adherence. Information on morbidity on the use of oral nutritional supplements will be collected using the questionnaire for gastrointestinal problems adapted from Pereira et al. (2014).

Baseline and 12 weeks A qualified phlebotomist will take a 5ml venous blood for analyses of the following: serum albumin, serum ferritin, serum cholesterol and haemoglobin

The intervention A 7-day ration of the oral nutritional supplement will be distributed to the intervention arm once a week. A day's serving of 330 ml is to be consumed in between meals. After endline data collection, the control arm will be given a ration of the oral nutritional supplement that will be enough for one month.

Statistical analyses

• To investigate the effect of sociodemographic characteristics, dietary intake and level of physical activity on the outcome measures, binary logistic regression for categorical variables, multinomial or ordinal logistics regression for multiple variables and linear regression for continuous variables
• To determine the efficacy of the oral nutritional supplement on outcome measures within the same treatment arm, repeated measures ANOVA and paired-sample t-test will be used
• To determine the differences in outcomes between the treatment groups the multivariate ANOVA and independent sample t-test will be used.

Data entry and management Field investigators will enter data collected on pre-programmed tablets that are functioning well. After each day's work, the supervisor will cross check and review all data in the presence of field investigators to ensure that complete and consistent data is finally submitted. The supervisor will flag any aberrant values and consult with the relevant field worker within 24 hours so that gaps will be filled, errors resolved and repeat interviews/visits can be conducted if need be. The submitted data will be accessed by the principal investigator daily for real-time review and monitoring. Once all data have been entered, reviewed and cleaned, final SPSS datasets will be generated and made available to the principal investigator for analysis. As soon as the manuscripts summarizing the study's primary outcomes are published, all datasets will be made available to any one who request it for research purposes.

Informed consent forms will be stored under lock and key.