**1. Introduction**

The ageing process in humans encompasses many changes that ultimately lead to undesirable clinical conditions like a higher fat deposition, osteoporosis, sarcopenia, frailty, and physical disability [1,2]. Current research on ageing focuses on decreasing the risk of developing the previously mentioned conditions and designing e ffective interventions to improve the patient's health [3–5].

Regarding sarcopenia, frailty, and physical disability, protein intake is one of the several factors linked to these clinical conditions [6]. A growing body of evidence suggests that daily protein intake above the recommended dietary allowance (R.D.A.) of 0.8 g/kg/d is associated with better physical performance, maintenance or even an increase in muscle mass, and decreased risk of physical disability. Therefore, it has been suggested to set the protein recommendation to a higher dose of 1.0–1.2 g/kg/d [7–10].

Additionally, it seems that other traits, like protein intake per meal and protein distribution, should be considered along with daily protein intake [11,12]. Nonetheless, the evidence is equivocal; other studies sugges<sup>t</sup> that the protein intake pattern might not be an important variable to consider in terms of skeletal muscle-related outcomes (e.g., strength and functionality) [12–14]. The studies supporting the importance of protein distribution in older adults sugges<sup>t</sup> that the consumption of 30 g of protein per meal or 0.4 g/kg per meal are associated with a higher skeletal muscle mass and strength and maximally stimulates muscle protein synthesis, respectively [15,16]. In this regard, some studies suggested that inadequate protein intake (<30 g/meal or <0.4 g/kg/meal) at specific meals might be a risk factor to consider as it is associated to a lower skeletal muscle mass, muscle strength, and functionality [17–19]. On the other hand, the number of meals that reach a protein content ≥30 g/meal or ≥0.4 g/kg/meal appears to be a protective factor as they are associated with a higher skeletal muscle mass, muscle strength, and lower physical disability [15,20,21]. Therefore, determining the percentage of older adults for both indicators might help visualize the magnitude of these two possible risk and protective factors.

Several studies have focused their attention on analyzing dietary protein intake patterns in older adults in di fferent countries [14,19–27]. Nonetheless, to the best of our knowledge, direct comparisons of protein intake patterns among countries are lacking [28]. Similarly, the comparison of the percentage of older adults that did not eat enough protein per day (i.e., 0.8, 1.0, and 1.2 g/kg/d) or per meal (30 g/meal; 0.4 g/kg/meal) among countries is missing. These comparisons among countries might serve as a starting point to understand the magnitude of this situation, because comparisons would help us to determine if these di fferent samples share common issues, allowing us to identify protein-eating patterns to be improved.

Therefore, this exploratory study aimed to (1) compare dietary protein intake patterns among older adults from four countries, (2) report and compare data of inadequate protein intake per day and per meal among older adults from four countries, and (3) analyze if these comparisons would yield similar results when the analysis was separated by sex. We hypothesized that dietary protein intake patterns would be di fferent, but inadequate protein intake would be similar among countries. Likewise, we hypothesized that the comparisons among countries separated by sex would yield di fferent patterns (e.g., if protein at breakfast di ffers among countries in women but not in men) and that the whole sample pattern would be the same in women but not in men.

#### **2. Materials and Methods**

#### *2.1. Study Design and Data Acquisition*

This is an exploratory analysis carried out with data from previously published articles where authors reported protein intake per day and meal in adults aged ≥60 years. Corresponding authors were contacted to gather demographic and protein intake data. We acquired data from two studies of two countries (Germany [14] and the United Kingdom [U.K.] [24]), from the National Health and Nutrition Examination Survey (NHANES) 2015–2016 publicly available database representing the United States of America (U.S.A.) [29], and data from our previous work in Mexico [18,20], all with cross-sectional designs.

To analyze the NHANES sample, we included data from participants with the following characteristics: (1) aged ≥60 year; (2) they were born in the U.S.A.; (3) reported an energy intake ≥600 and ≤4000 kcal/day; and (4) had complete data for age, height, and body mass. From these records (*n* = 1039, 50% women), we randomly selected 200 subjects (100 per sex to keep the sex proportion) to decrease the differences in sample size among groups. There were no significant differences between included and nonincluded subjects (*n* = 839) for age (*p* = 0.81), body mass (*p* = 0.55), height (*p* = 0.38), BMI (*p* = 0.98), nor total protein intake per day (*p* = 0.15). This sample was not weighted according to the NHANES complex study design as the other studies did not follow the same sampling design.

All studies independently coded the three main meals (i.e., breakfast, lunch, and dinner), and reported that they obtained participants' written informed consent and ethical approval from their local institution before any assessment. Table 1 shows an overview of the included samples.

#### *2.2. Protein Intake Variables*

When studies reported two or more days of dietary assessment, we averaged the protein intake per day and per meal, and we used these averages for further analysis. We calculated relative protein intake per day (g/kg body mass/d) and per meal (g/kg body mass/meal), meal contribution to total daily protein (%), and protein distribution coefficient of variation in addition to the absolute protein intake per day (g/d) and per meal (g/meal).

*Meal contribution* to total protein was calculated as:

$$\text{Mean contribution} \quad = \text{ PM/TP} \times 100,\tag{1}$$

where *PM* is the protein reported for any given meal (g) and *TP* is the total daily protein intake (g). Theproteindistributioncoefficientofvariation(*PDCV*)wascalculatedas:

*PDCV* = *SDP*/*MP*, (2)

where *SDP* is the standard deviation of the three main meals and *MP* is the mean protein intake for the three main meals.

#### *2.3. Inadequate Protein Intake*

Inadequate protein intake (IPI) was considered as any protein consumption <0.8 (IPID-0.8), <1.0 (IPID-1.0), and<1.2 (IPID-1.2) g/kg/d [7,8] or<30 g/meal (IPIM-30) and<0.4 g/kg/meal (IPIM-0.4) [15,16]. We reported IPIM-30 and IPIM-0.4 for each main meal. We also counted the number of meals per day (coded as zero [0M], one [1M], and two or three meals [+2M]) with ≥30 g protein and ≥0.4 g protein/kg each.



NHANES: National Health and Nutrition Examination Survey; W/M: number of women and men. ‡ The complex sampling design of NHANES leads to nationally representative data; however, this is not the case for this study as the data were not weighted according to its sampling design and were composed of a smaller sample size (200 vs. 1039).
