The (Cost)Effectiveness of Increasing Protein Intake on Physical Funtioning in Older Adults

In this RCT with the duration of 6 months the investigators will examine the long term (cost) effectiveness of increasing protein intake to at least 1.2 g/kg adjusted body weight/d on physical functioning in older adults with habitual low protein intake. Additionally, the investigators will examine the combined effect of increasing protein intake to at least 1.2 g/kg adjusted body weight/d and consuming protein in close proximity with regular physical activity on physical functioning in older adults with habitual low protein intake.

Three sub-studies will be conducted, of which the main objectives are to:

• Examine the effect of persuasive technology on adherence to consumption of protein rich food products in order to increase protein intake to at least 1.2 g/kg adjusted body weight/d, and to the combination of increasing protein intake to at least 1.2 g/kg adjusted body weight/d and consuming protein in close proximity with regular physical activity.
• Examine the effect of increasing protein intake to at least 1.2 g/kg adjusted body weight/d on faecal and oral microbiota composition in older adults with a habitual low protein intake.
• Examine the effects of increasing protein intake to at least 1.2 g/kg adjusted body weight/d on food-stimuli related central nervous system satiety and reward responses involved in the regulation of food intake, measured by functional magnetic resonance imaging in older adults with a habitual low protein intake.

Study design: Randomized controlled trial with the duration of 6 months in two study sites: Amsterdam, the Netherlands and Helsinki, Finland. Stratification by gender and habitual protein intake (low protein (>=0.9 g/kg BW/day - <1.0 g/kg BW/day), very low protein (<0.9 g/kg BW/day)).

Study population: A total of 264 community-dwelling older adults aged ≥ 65 years with an habitual low protein intake (n=132 at each study site).

Intervention: This RCT consists of three groups; two intervention groups and one control group. Intervention group 1 (N=44 at each study site) will receive personalized dietary advice aimed at increasing protein intake to at least 1.2 g/kg adjusted body weight/d without changing daily energy intake, by regular foods and by provided protein-enriched food products. Intervention group 2 (N=44 at each study site) receives personalized dietary advice similar to group 1 and also receives personalized advice to consume protein rich foods in close proximity of usual physical activity. All groups receive a standard brochure of the Netherlands or Finnish Nutrition Centre with general information about healthy eating habits. The control group (N=44 at each at each study site) receives no further intervention.

Main study parameters/endpoints: The primary outcome of this study is change in walk time on the 400 meter walk test. Secondary outcomes are change in dietary intake (including macro- and micronutrients), malnutrition prevalence, physical performance, mobility limitations, muscle strength, body weight and body composition, frailty status, quality of life, and health care costs.

Statistical analyses:

The collected data at the two study sites will be pooled together.

As a result of randomization at study baseline, we assume that groups are equal regarding demographical and socio-economic variables. If this is not the case, we will adjust for differences between groups at baseline. We will adjust for study site (the Netherlands, Finland) and baseline outcome values. We will present unadjusted and adjusted results.

The main analyses will be based on intention-to-treat principles, but per-protocol analyses will also be conducted as a sensitivity analysis.

Multiple Imputation (MI) using multivariate Imputation by Chained Equations (MICE) will be used to impute missing cost and effect data. For this, the missing values need to be missing at random.

The (cost) effectiveness of two intervention groups will be examined against the control group on the primary outcome walk time on the 400 meter walk test. We will compare outcomes between the respective intervention groups and control group separately to determine whether the two interventions are effective (group 1 versus control group; group 2 versus control group).

We will perform mixed model regression analyses adjusting for confounding variables at baseline and study site as cluster variable. We will not correct for multiple testing but look at the clinical relevance of the outcome.

In addition, we will perform sensitivity analyses leaving out the participants who took part in the persuasive technology sub-study.

**** Update June 2020 ****

Deviation of the protocol

Due to the worldwide spread of Covid-19, the original protocol of the PROMISS prevention trial has been changed for those participants who were still active in the study during the spread (March 16th - June 1st). In consultation with both Medical Ethical boards from Finland and the Netherlands, and PROMISS' ethical advisor, the following changes were applied:

March 2020 * Starting from March 16 2020, the final follow-up measurement of 80 participants (out of 276) was postponed until further notice.

• Participants were informed that they would be invited for the final follow-up measurement when the country-specific governmental regulations allowed it.

• Participants were requested to stick to their habitual diet (control group) or their intervention diet (both intervention groups).

April 2020 * Final measurements were resumed (through interview by phone) except for the physical measurements.

• Dietary intake was assessed in the week prior to the phone call measurement.

• Self-reported body weight was added to the questionnaire.

• The physical follow-up measurement was still postponed until further notice.

May 2020

• Data collection was continued by means of questionnaires during the phone call measurement in both Finland and the Netherlands.

June 2020

• Starting from the beginning of June, the physical measurements at the clinic site were also resumed in both Finland and the Netherlands. Those with no health complaints potentially caused by the coronavirus were allowed to visit the clinic site.
• The data collection will finish by July 31, 2020.

We will perform sensitivity analyses excluding those participants who were still active in the study during the spread and thus had their physical follow-up measurement +/- 8 months after the baseline assessment (instead of 6 months).

Inclusion criteria

• Age ≥ 65 years;
• Community-dwelling;
• Lower protein intake defined as both a probability score above a certain cutoff on the protein screener (www.proteinscreener.nl) as well as based on actual protein intake assessed by 24-hour recalls. The protein screener was developed and validated using an extended FFQ among Dutch older adults. The cutoff will be chosen based on results of different studies in which the investigators compare the probability scores of the protein screener with protein intake as measured with food diaries and/or dietary recalls. The investigators will then choose the probability score that is most closely associated with a protein intake < 1.0 g/kg adjusted body weight/day. This probably score reflects older adults with a higher probability on a protein intake < 1.0 g/kg adjusted body weight/d than a general sample of older adults;

Exclusion criteria

• Inability or unwillingness to provide informed consent
• Not able to eat independently;
• Not able to speak, write and read the Dutch language;
• Current participation to supervised behavioral or lifestyle intervention that intervenes with PROMISS intervention;
• Not able the visit the research site in the following next 6 months;
• Bedridden or wheelchair bound;
• Individuals who do not go outside;
• Diagnosed with severe kidney disease;
• Diagnosed with Parkinson's disease;
• Diagnosed with diabetes mellitus type I;
• Diagnosed with diabetes mellitus type 2 and starting with insulin;
• Current treatment of cancer (with the exception of basal cell carcinoma);
• Vegan diet;
• Severe allergies to certain food products (such as peanuts, gluten);
• Diagnosed with an eating disorder (self-reported);
• Purposefully lost/gained > 3 kg in the past three months
• Heart problems in the past three months (heart attack, angioplasty, heart surgery, stroke or other serious heart disease)
• Not able to complete the 400 meter walk test within 15 minutes (self-reported, and assessed at study baseline).
• Alcohol abuse past 6 months (AUDIT-C ≥ 2);
• Low cognitive status, defined as the mini-mental state examination (MMSE) score ≤ 20
• BMI < 18.5 kg/m2 (self-reported, and assessed at study baseline);
• Overweight, defined as BMI > 32.0 kg/m2 (self-reported, and assessed at study baseline);