Measurements

This section describes each measurement performed to evaluate the longitudinal effects of the chosen interventions. Table 3 displays the time points for each examination. At baseline, we test the peak isometric and isokinetic peak torques as well as leg extensor power in both legs to evaluate if a large side difference between thigh peak torques is associated with an improved or diminished response to the interventions.

Time schedule of participant engagement

Abbreviations: MRI CSA magnetic resonance imaging cross-sectional area, DXA dual-energy X-ray absorptiometry, FSR fractional synthesis rate, PSQI Pittsburgh Sleep Quality Index, SF-36 36-item Short Form Health Survey

A change in muscle CSA is a direct reflection of the intervention’s ability to restore muscle protein and a direct consequence of the net protein balance over time in the target muscles. We chose m. quadriceps and m. vastus lateralis CSA based on MRI scans of the dominant thigh as the primary outcome, as it directly reflects the intervention’s ability to counteract age-related loss of muscle mass. MRI scans have high sensitivity and validity in terms of lean mass in skeletal muscle [61], and previous researchers have considered this modality superior to DXA [62, 63].

Each scan consists of six axial slices, with the first slice placed in the medial tibia plateau. Each slide is 8 mm thick and separated by a 60-mm gap as shown in Fig. ​Fig.4.4. The primary time interval for assessment is from baseline to 12 months.

Magnetic resonance imaging analysis. We place slices as shown for analysis of cross-sectional area of the m. quadriceps femoris muscle and analyze slices 3 (counting in distal to proximal direction) and 4 for all subjects, and we use slice 4 for primary outcome evaluation. We fix the placement of slices in absolute distances, but we measure the femur length on dual-energy X-ray absorptiometric scans from the lateral tibial plateau (0 %) to the top of the greater trochanter (100 %) to report the relative placement of slices. Currently, placement of slice 3 ranges from 27 % to 36 % and slice 4 from 40 % to 54 % of the femoral length, depending on the height of the participant

After having the participant perform a brief warm-up on a cycle ergometer, a tester will measure the dominant thigh peak strength at 70-degree flexion (0 degrees represents horizontal) in a Kinetic Communicator (model 500-11, Kinetic Communicator; Isokinetic International, Chattanooga, TN, USA) with verbal encouragement. An investigator will choose the best of three sweeps (i.e., highest peak force) for analysis. The rate of force development and impulse will also be analyzed [51, 64].

Using the same settings as those used for the isometric peak strength exercise, the same tester will measure the dominant thigh isokinetic (at 60 degrees/second) force. With verbal encouragement, participants will perform maximal sweeps until peak values decline markedly from the best sweep on that particular test day, and the test is finished. An investigator will choose the best sweep (i.e., highest peak torque) and analyze the peak torque, total work, and angle at peak torque [51, 64].

After recording the isometric and isokinetic peak torques, the tester will measure the maximum single-leg extensor power of the dominant leg using a University of Nottingham leg extensor power rig according to procedures described elsewhere [65]. This test is known to correlate with physical performance [8] even more than muscle strength [66] and serves to link the changes in strength to functional outcomes. During the test, participants are in a seated position, and a single maximum explosive leg extension accelerates a flywheel from rest. A computer calculates the power of the leg extensors on the basis of the speed of the flywheel. The participants familiarize themselves with the procedure by performing two warm-up trials followed by a minimum of five and a maximum of ten maximal trials with approximately 30 seconds of rest between them.

The 30-second chair-stand test is a functional assessment of strength and endurance in the lower extremity by measuring the number of stands completed in 30 seconds with hands crossed on the chest [67]. The test is useful in characterizing the participant’s functionality and is also known to correlate with daily activity level [68].

Working in sterile conditions, a qualified researcher will obtain muscle biopsies from the dominant m. vastus lateralis ad modum Bergström [69] with suction using the local anesthetic lidocaine 1 %. One biopsy is taken before and one after 12 months of the intervention, approximately 2–3 cm apart in a proximal-to-distal direction.

A trained operator will perform whole-body DXA using the enCORE v.16 software (Lunar iDXA; GE Medical Systems, Pewaukee, WI, USA). Participants arrive having refrained from solid foods from 21:00 the day before the baseline and 12-month scans, and we perform scanning between 08:00 and 10:00. We obtain the remaining scans in the fed state at all times of the day, but prior to all scans participants are euhydrated and instructed to void. Whole-body composition is autoanalyzed, and, on a separate occasion, a blinded investigator performs separate thigh analyses manually with a region of interest (ROI) defined proximally by drawing a horizontal line laterally from the distal part of the groin and distally by drawing a line horizontally through the medial tibial plateau [70–72]. The thigh ROIs are equal in size in each subject but vary with thigh length between subjects.

In the same baseline and 12-month DXA sessions, the same operator scans the bone mineral density in the lumbar region (L2–L4) and the dominant collum femoris. The investigator will place the vertebral ROI as caudal as possible, including the discus caudal to the vertebra, making sure not to include the adjacent vertebra or spinous process. On the collum femoris scan, the investigator places the ROI as distal as possible without including the greater trochanter [73, 74].

At 0, 6, and 12 months, participants place a fecal sample in an insulated bag with freezer elements until delivery at Bispebjerg Hospital within 48 h. The container is then stored at −60 °C until further analysis. Following homogenization, we extract the total DNA using the PowerSoil DNA isolation kit (MO BIO Laboratories, Carlsbad, CA, USA) with an initial bead-beating step and employ tag-encoded 16S ribosomal RNA gene (prokaryotes) and internal transcribed spacer region (eukaryotes) high-throughput sequencing using the MiSeq platform (Illumina, San Diego, CA, USA) to characterize the prokaryotic and eukaryotic components of the gut microbiome. As previously reported [75], we purify, extract, and sequence the virus-like particles for characterization of the gut virome.

We will analyze the fecal metabolite extracts by performing gas chromatography with time-of-flight mass spectrometry (GC-TOF-MS) and nuclear magnetic resonance (NMR) spectroscopy using the same metabolite extract. The extraction procedure will involve stabilization of the homogenized fecal samples in PBS (pH 7.4) followed by freeze-drying and subsequent resuspension in methanol. We will perform untargeted metabolomics analysis and targeted profiling of short-chain fatty acids with a high-throughput GC-TOF-MS setup consisting of an Agilent 7890B gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) and a high-throughput Pegasus TOF-MS spectrophotometer (LECO, St. Joseph, MI, USA). To increase GC-TOF-MS sensitivity toward a broad spectrum of nonvolatile metabolites, we will derivatize the samples as previously described [76] followed by processing of the obtained complex raw data using state-of-the-art three-way decomposition methods with Parallel Factor Analysis 2 [77]. We will further stabilize the fecal metabolite extracts using PBS prior to performing one-dimensional ¹H NMR measurements using an AVANCE III 600 spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany) operating at a Larmor frequency of 600.13 MHz for protons and equipped with a cooled high-capacity autosampler (SampleJet; Bruker BioSpin GmbH) [78].

We will collect the plasma samples with participants in fasting condition prior to the oral glucose tolerance test (OGTT) at 0 and 12 months and in fed condition at 6 and 18 months. Researchers will draw the blood into two containers, one containing tripotassium ethylenediaminetetraacetic acid (K₃-EDTA) and one with heparin, and, after centrifugation at 3970 rpm for 10 minutes at 4 °C, we will pipette plasma into vials that are stored at −60 °C. We then will perform untargeted metabolomics of all samples using proton one-dimensional ¹H NMR spectroscopy and for a subset of the samples by GC-TOF-MS.

We will perform the NMR measurements according to standard operating procedures developed for NMR metabolomics of plasma samples [79], and the data will provide unbiased metabolic fingerprints of the plasma samples, including information about lipoprotein particle distribution [80]. Prior to GC-TOF-MS analysis, we will thaw, vortex, and centrifuge the plasma samples at room temperature followed by addition of ice-cold acetonitrile in 1:3 (vol:vol) to precipitate proteins. Immediately after this, we will vortex the samples vigorously and centrifuge them at 20,000 × g for 10 minutes at 4 °C, and we will then dry 50 μl of clear supernatant under reduced pressure, derivatize it as described previously [76], and subject it to GC-TOF-MS.

To assess MPS, we will perform an 8.5-h primed, continuous infusion of a stable isotope-labeled amino acid tracer, L-[ring-¹³C₆]phenylalanine (Cambridge Isotope Laboratories, Tewksbury, MA, USA) at baseline and after 12 months of the intervention in a random subset of participants from each of the five intervention groups. The participants will fast overnight fast, and we will start the tracer infusion followed by collection of muscle biopsies after 1.5-h and 4.5-h infusions to measure the basal resting MPS over a 3-h period. Thereafter, we provide a protein drink as a beverage containing 20 g of whey hydrolysate plus 10 g of maltodextrin followed by collection of the third muscle biopsy. After an additional 4 h, we will measure the postprandial response of MPS. We will sample the muscle biopsies from the m. vastus lateralis muscle approximately 3 cm apart in a proximal-to-distal direction. Throughout the day, we will collect venous blood samples to measure tracer enrichment as well as amino acid and insulin concentrations, and we will measure the enrichment of phenylalanine tracer in blood and free in muscle cells by GC-MS/MS, whereas we will measure the incorporation of labeled phenylalanine in muscle proteins using GC-combustion/isotope ratio MS. We will then calculate the fractional synthesis rate (FSR) as the difference of incorporated tracer between subsequent biopsies (E_p2 − E_p1) divided by the precursor enrichment (E_precursor; venous plasma and muscle free enrichment) and divided by incorporation time: FSR (%/h) = [(E_p2 − E_p1) × (E_precursor)⁻¹ × (incorporation time)⁻¹] × 100 % [81]. This setup allows the evaluation of both basal MPS and the MPS response to a protein/carbohydrate drink before and after the 12-month intervention as a measure of muscle anabolic responsiveness to intake of a single protein supplement. We will also analyze muscle biopsies for relevant myocellular gene expression using real-time reverse transcriptase-polymerase chain reaction and for protein signaling through, for example, mammalian/mechanistic target of rapamycin complex 1 by performing Western blotting.

At baseline, we will obtain standard health screening blood samples. At the remaining time points, we will measure only plasma HbA1c, cholesterol, and creatinine. Moreover, we will measure body weight, waist, and hip circumference at the remaining time points, whereas we will measure only height at baseline. Participants will wear underwear only when we measure body weight at baseline and 12 months, and they will wear light clothing at 6 and 18 months.

We will evaluate hand grip strength of the dominant hand (bilateral testing at baseline) by using a hand grip strength dynamometer (DHD-1 [SH1001]; SAEHAN Corporation, Changwon City, South Korea). The tester allows at least 30 seconds of rest between a minimum of three attempts, and the highest value at the given time point is used. The test is finished when two consecutive measurements are lower than the peak value [82].

On a 20-m course marked with two colored cones, the physical examiner instructs the subject to walk 400 m as fast as possible without personal support or sitting down. Up to 1 minute of standing is permitted if the participant feels tired or experiences discomfort, as long as the test is completed within 15 minutes [83]. Among several functional measures, we chose 400-m gait speed as it is one of the more demanding functional tests available for elderly populations. This minimizes the risk of a ceiling effect among the included participants whom we expected to be relatively well-functioning.

At baseline and 12 months, we will hand out Danish translations of the 36-item Short Form Health Survey [84] and the Pittsburgh Sleep Quality Index [85], which the subjects complete without further instructions.

We measure 4-day (96-h) overall activity by mounting an activity monitor (activPal 3™, activPal 3c™, or activPal micro; PAL Technologies, Glasgow, UK) on the anterior surface of the thigh. The period always covers an entire weekend. An investigator will analyze data quantified as step counts, time in sitting/lying position, time in standing position, and time walking [86, 87].

We will administer a 3-day weighed food and liquid registration from Wednesday to Friday before, 4–6 weeks after, and 50 weeks after commencing the intervention. Investigators subsequently will quantify total daily intake and meal distribution of macronutrients using the MADLOG VITA system (MADLOG ApS, Kolding, Denmark).

Participants will complete a questionnaire about their experience with taking the dietary supplements. We will collect baseline data on the first consumption day (day 0) and weekly for 12 weeks. Thereafter, we will reevaluate participants every third month until week 52. The flavor of the supplements will be alternated every week between fruit and cacao, and the participants will fill in the questionnaires within 1 minute after ingestion of the supplement. We will measure the participants’ acceptance after the first sip and after complete consumption on a 9-point hedonic scale ranging from 1 = “do not like at all” to 9 = “like a lot,” with neutral in the middle. After the first sip, we will record the participants’ perceived experiences using the check-all-that-apply method with 20 attributes of taste, flavor, mouth feel, and sensory appeal of the dietary supplement. After participants consume the supplements, we will collect data on how satiating, refreshing, and easy to drink it was, as well as on the strength and persistence of the aftertaste on a 9-point scale ranging from 1 = “completely disagree” to 9 = “completely agree,” with neutral in the middle.

Subjects will complete the Satisfaction with Food-related Life scale [88] on days 0, 90, 180, and 360 to identify factors contributing to satisfaction with food-related lifestyle.

Participants will undergo two OGTTs during the study. Participants will fast overnight, and a researcher will draw a basal venous blood sample and then administer 75 g of anhydrous glucose dissolved in 250 ml of tap water. Subsequently, the researcher will draw blood at 45 and 120 minutes after consumption of the glucose. Participants will lie supine throughout the 2-h period [89]. We will draw blood in K₃-EDTA vials and cool the vials for at least 15 minutes, centrifuge the sample for 10 minutes at 3970 rpm at 4 °C, immediately analyze the plasma for glucose, and store aliquots at −80 °C for subsequent insulin measurement. At baseline, we draw two further blood samples in K₃-EDTA and heparin vials respectively, which are cooled on ice, centrifuged for 10 minutes at 3172 × g, and the plasma is then stored at −60 °C until metabolome analysis.

Upon participant inclusion, a researcher will conduct a short, standardized, qualitative interview with the participant to gain information on marital status, living conditions, work life, hobbies, and dietary preferences. At the 18-month follow-up, a researcher will interview the participants to gain information about their experience with the intervention and current habits regarding nutrition and exercise.

At baseline, the staff will hand out a questionnaire with a range of questions about the food perceptions and habits of the participant. The questionnaire combines basic socioeducational data, quantitative questions, and quantifiable qualitative questions about lifestyle changes, dietary changes and perceptions, and intake of protein-rich foods. The participants can complete the questionnaires during the first 3 months and subsequently hand them in.

We will select 25 of the intervention participants for 1.5- to 2.5-h life story/trajectory interviews in their own homes. The main themes of the interviews are past and present perceptions of and habits concerning food and physical activity, but we will also include questions regarding mobility and work life.

We will select 48 research participants to engage in a qualitative study focused on everyday life routines, eating practices, physical activity, and experiences with participating in the trial. We do this by observing the research participants in their homes and at the site of the clinical trial, thereby following their daily activities in both settings. Further, we will conduct semistructured qualitative interviews to gain in-depth knowledge about selected individuals.