Protein enrichment of familiar foods as an innovative strategy to increase protein intake in institutionalized elderly

Abstract

Objective

To increase the protein intake of older adults, protein enrichment of familiar foods and drinks might be an effective and attractive alternative for oral nutritional supplements (ONS). We performed a pilot study to test whether these products could help institutionalized elderly to reach a protein intake of 1.2 gram per kg body weight per day (g/kg/d).

Design

Intervention study with one treatment group (no control group). Dietary assessment was done before and at the end of a 10-day intervention.

Setting

Two care facilities in Gelderland, the Netherlands: a residential care home and a rehabilitation center.

Participants

22 elderly subjects (13 women, 9 men; mean age 83.0±9.4 years).

Intervention

We used a variety of newly developed protein enriched regular foods and drinks, including bread, soups, fruit juices, and instant mashed potatoes.

Measurements

Dietary intake was assessed on two consecutive days before and at the end of the intervention, using food records filled out by research assistants. Energy and macronutrient intake was calculated using the 2013 Dutch food composition database. Changes in protein intake were evaluated using paired t-tests.

Results

Protein intake increased by 11.8 g/d (P=0.003); from 0.96 to 1.14 g/kg/d (P=0.002). This increase is comparable to protein provided by one standard portion of ONS. The intake of energy and other macronutrients did not change significantly. At the end of the intervention more elderly reached a protein intake level of 1.2 g/kg/d than before (9 vs 4). Protein intake significantly increased during breakfast (+3.7 g) and during the evening (+2.2 g).

Conclusion

Including familiar protein enriched foods and drinks in the menu helped to meet protein recommendations in institutionalized elderly.

Introduction

Many elderly in the Netherlands experience undernutrition. In their report about undernutrition, the Dutch Health Council reports that 33% of the hospitalized elderly, 18-21% of the institutionalized elderly and 12-16% of the elderly receiving home care suffer from undernutrition (1). Undernutrition is caused by an inadequate intake of calories, protein, or other nutrients needed for tissue maintenance or repair (2).

The reasons for an inadequate nutrient intake in the elderly are diverse, including anorexia of ageing, changes in food preferences and difficulties in obtaining and preparing food (3., 4., 5.). Moreover, elderly people report to have declined appetite feelings (6) and eating meals is no longer a desire but a discipline for many elderly (7). The elderly especially have an increased risk at an insufficient protein intake because preferential consumption of protein-rich foods may decrease with ageing (5).

They have, however, also an increased need for proteins due to a reduced ability to use available protein (8). The current recommended protein intake for elderly over 65 years is the same as for younger adults: 0.8 gram of protein/kg body weight/day (g/kg/d). Many researchers and geriatricians plea for a higher recommended protein intake: 1.2 - 1.5 g/kg/d. They argue that this higher recommendation is not just to prevent deficiencies but also to maintain health and function in the elderly (8., 9., 10., 11., 12., 13.). Moreover, recent literature suggests that an intake of 25-30 gram of high quality protein per meal is needed to maximize muscle protein synthesis and maintain muscle mass in elderly people (11). Recent studies suggest that most elderly do not reach this high intake with their current diet (14., 15., 16.). Especially breakfast of Dutch elderly is low in protein content (14).

Therefore, professionals responsible for nutritional care look for ways to improve protein intake in undernourished elderly. Normally first, they try to increase protein intake by advising protein rich snacks and double sandwich toppings, like cheese or meat. If this is not successful, protein enriched oral nutritional supplements (ONS) are commonly prescribed. However, the effectiveness of ONS on functional improvements is still a matter of debate (17). When ONS is consumed for prolonged periods, the compliance usually declines; the number of different flavors and textures is too limited to fulfil the elderly's needs and wishes (18., 19., 20.).

To fill the gap between regular foods and clinical nutrition, we decided to develop protein enriched food products. As part of the product development phase, we interviewed undernourished elderly to gain consumer insights. We found that the elderly prefer familiar foods that are easy to consume and prepare and that portion sizes should not increase. By enriching foods and drinks that are familiar to the elderly, they can increase their protein consumption without changing their eating habits or increasing their portion sizes. Besides these advantages, these protein enriched familiar products are ready to eat in contrast to commonly used protein powders that have to be added to foods with the risk of decreasing palatability. Moreover, the newly developed products were tested in a consumer panel of healthy elderly who rated them as more palatable than ONS.

If protein enriched products are consumed in the same amount as regular products, protein intake will automatically increase. However, protein enriched products are suggested to be more satiating than iso-caloric carbohydrate enriched products (21). This might limit the intake of protein enriched products. Another undesired effect could be that the higher protein intake from the enriched products is compensated by choosing low protein foods and drinks during the rest of the day.

Therefore, this pilot study investigated whether protein enriched familiar food products, specifically developed for older adults, enabled them to reach a protein intake of 1.2 g/kg/d. Furthermore, the protein distribution across meals will be assessed. The results of this pilot study will be used to improve products for an intervention study on the health effects of a protein enriched diet in elderly patients.

Methods

Design

This pilot study was performed in the Netherlands in two care facilities from the same care organization. The first facility was a residential care home where elderly people live long-term, and the second facility was a rehabilitation center for temporary stay. A total of 88 elderly resided in the care home, while the rehabilitation center had room for 32 people at once. People in the rehabilitation center were for instance recovering from surgery or stroke. Study participants were recruited from the residents in three ways: by personal information brochures in their mailboxes, by posters in the common areas of the care facilities, and by the nursing staff who asked them if they wanted to participate. Residents had to give written informed consent prior to receiving the nutritional intervention. This pilot study consisted of one intervention group, without having a separate control group. The Wageningen University Medical Ethical Committee approved this study. This study was registered on ClinicalTrials.gov (Identifier: NCT02141256).

Participants

Participants were at least 60 years of age and stayed at one of the care homes. Potential participants were not included when they were cognitively impaired; suffered from dementia; had dysphagia; only received tube feeding; or had dietary protein restrictions due to for example chronic kidney disease or food allergies.

Nutritional intervention

During a 10-day period a variety of newly developed food products were incorporated into the food assortment of the care homes. These included protein enriched foods (e.g. bread, soups, fruit juices, and mashed potatoes) and foods with a naturally high protein content (e.g. veal). Some of these foods could be used as a replacement of comparable regular foods, such as the bread, soups, mashed potatoes and meat. Other foods were offered as additional choices, such as the protein enriched fruit juices. Most intervention products were offered as an extra option within the regular menu, because we wanted to test whether these products were effective when subjects had free choice, which reflects daily practice. Table 1 shows all intervention products, with the variety of flavors and their additional protein content per portion. The protein enriched products were enriched with protein from plant and/or animal origin, e.g. soy and dairy. Participants were completely free to consume the protein enriched products or not, and could consume as much of the foods as they wanted.

Table 1.

Protein enriched intervention products and their additional protein content per portion compared to non-enriched products

Product	Variety of flavor or type	Additional protein per portion (gram)	Portion size
Bread	Light	5.6	2 slices (27 gram per slice)
	Dark	5.6	2 slices (27 gram per slice)
Soup (without meat)	Mushroom	10	150 mL
	Broccoli	10	150 mL
	Tomato	10	150 mL
Fruit juice	Orange	10	150 mL
	StrawberryApple	10	150 mL
	Blue berry-Apple	10	150 mL
Mashed potatoes	-	8.4	150 gram

The food distribution system differed between the two facilities: the elderly in the care home consumed only their hot meal in the restaurant, while the elderly in the rehabilitation center consumed all three of their meals in the center's restaurant. Juices and snacks were placed in participants' refrigerators, so they could choose themselves if and when to take them.

Outcome measurements

At baseline, a participant's descriptive measures were recorded, including birth date, gender, body weight, height, and risk of malnutrition. Body weight was not expected to change in this short time period and was therefore measured once during the study with a calibrated digital weighing scale to 0.01 kg (SECA weighing scale). When a participant could not stand on the weighing scale, the nursing staff reported body weight that was recently measured with the weighing chair. Body weight was also recorded to calculate the protein intake in g/kg/d. Height was measured to 0.1 cm using a stadiometer (SECA stadiometer). When a participant could not stand up straight, due to physical restraints, lower leg length was measured to 0.1 cm and the formulas of Sienkiewicz-Sizer (22) were used to estimate standing height. These formulas were developed specifically for elderly people. One researcher screened for risk of malnutrition with the Mini Nutritional Assessment Short Form (MNA-SF).

Dietary intake was assessed twice: on two consecutive days a week before the intervention period started, and on the intervention's last two days (days 9 and 10). Three trained research assistants filled out food records with the participants. To prevent inter-rater variability the same research assistant visited the same participant each time. The assistants visited the participants on average 3 times a day to ask the participants what they consumed. This assessment method was chosen to limit recall bias. For the hot meal the procedure was slightly different: research assistants were present during the meal. When a plate was served, each component of the meal was recorded in household measures (amount of spoons or portion sizes). Directly after the hot meal, the participant was asked how much of the served plate was left over.

Consumption of the intervention products (at the end of the intervention) and their corresponding regular products (before the intervention) was calculated in portions per day. Furthermore, dietary intake of energy, macronutrients and food products was calculated with the program Compl-eat (Department of Human Nutrition, Wageningen University), using the 2013 Dutch food composition database (23). Protein intake was calculated in g/kg/d. This was done per kg actual body weight but also per kg adjusted body weight when BMI was above 27 kg/m2, because protein recommendations are based on lean body mass. Body weight was adjusted for subjects with a BMI > 27 kg/m2 to a body weight corresponding with a BMI of 27 kg/m2 (24). This is in line with the practical guidance for Dutch dieticians when calculating protein requirements. Protein intake in g/kg actual and adjusted body weight per day was compared with the reference intakes of 0.8 and 1.2 g/kg/d. Participants could also comment on a product's taste or texture during the whole intervention period. Comments were recorded and interpreted in a qualitative manner. Comments regarding non-food items such as comments on health were also recorded.

Sample size calculation

Sample size was calculated to detect an increase of 15 gram protein per day as statistically significant. With a SD of 25 gram a minimum of 22 subjects was required (power=0.80, α=0.05).

The used SD is slightly lower than the SD in a study conducted in a population of Dutch hospitalized elderly (15), because we expected that our subjects would have a more stable intake. To account for a dropout rate of 10% a sample size of 25 subjects was considered sufficient.

Statistical analysis

Statistical analysis was done using IBM SPSS Statistics Version 22. Descriptive statistics were performed to describe baseline characteristics. All continuous variables are presented as means ± SD. To investigate whether the effect on the protein intake was different in the two different facilities, we used the Univariate General Linear Model (GLM) procedure. The change in protein intake is analyzed using a paired samples T-test. The changed intake of energy and other macronutrients was also analyzed using a paired samples T-test. A P<0.05 was considered statistically significant.

Results

From the 88 residents in the care home, 21 were excluded based on their cognitive capabilities. From the 67 eligible people, only 11 wanted to participate. Reasons for not wanting to participate were: they were not interested, they thought it was too exhausting, or they did not want to participate without specifying a reason. In the rehabilitation center, all new guests were informed about the study. In total, 30 people have been informed, until we had 14 participants in this center. In total, 25 participants gave written consent to participate in the study: 11 from the residential care home and 14 from the rehabilitation center. From the 11 participants in the care home, one withdrew before the start of the intervention due to health problems. In the rehabilitation center, one participant was unexpectedly discharged before completion of the baseline measurements and one was critically ill during the last two measurement days. All three subjects were excluded from the statistical analyses, leaving data collected from 22 subjects for statistical analyses.

Table 2 shows the baseline characteristics of the study population. This study population had a mean age of 83.0 ± 9.4 years. According to the MNA-SF scores, 6 of the 22 participants were undernourished, and 4 were at risk of becoming undernourished.

Table 2.

Baseline characteristics of study population (n=22; 13 female and 9 male)

	Mean ± SD	Range
Age (years)	83.0 ± 9.4	61 - 95
Body weight (kg)	73.4 ± 17.6	46.0 - 116.5
Height (m)	1.64 ± 0.08	1.50 - 1.82
Body Mass Index	27.3 ± 6.0	17.0 - 40.6
MNA-SF score	n (%)
Normal nutritional status (12-14)	12 (54.5)
Risk of undernutrition (8-11)	4 (18.2)
Undernourished (0-7)	6 (27.3)

First, the consumption of bread, juice, soup, mashed potatoes, and meat was calculated in portions per day before and at the end of the intervention (Table 3). For the end measurement, the portions of intervention products were also calculated for these product groups. The consumption of bread increased a little, and at the end of the intervention 2 of the 3 slices bread that were consumed were protein enriched. The protein enriched bread delivered 10.8 g protein. The amount of juice increased due to the intervention juices from 0.5 to 1.2 portions, which increased protein intake with 7 g. Almost all protein delivered via juice, came from the protein enriched juices: 6.7 g. Consumption of soup remained stable, but almost half of it was protein enriched at the end of the intervention: 0.3 of the 0.8 portions in total. More than half of the protein intake from soup was delivered by the intervention soups: 5.3 g of 7.8 g. Mashed potatoes and meat were not consumed in large portions. This can be explained by the multiple choices that were available in the menus.

Table 3.

Consumed portions per day of bread, juice, soup, mashed potatoes and meat before and at the end of the intervention

	Before		Intervention
	Portions, mean (range)	Protein delivered (g)	Portions, mean (range)	Protein delivered (g)
Bread a	2.5 (0.5-4.5)	8.8	2.9 (1.1-5)	13.3
of which intervention bread products	-	-	1.9 (0-4)	10.8
Juice b	0.5 (0-3.2)	0.4	1.2 (0-4)	7.0
of which intervention juices	-	-	0.7 (0-3)	6.7
Soup c	0.7 (0-1.4)	4.4	0.8 (0-1.4)	7.8
of which intervention soups	-	-	0.3 (0-1)	5.3
Mashed potatoes	0.13 (0-1)	0.9	0.34 (0-1)	2.8
of which intervention mashed potatoes	-	-	0.16 (0-1)	1.7
Meat	0.73 (0-2)	12.8	0.69 (0-2)	14.2
of which intervention meat	-	-	0.16 (0-1)	2.9

^a1 portion is 1 slice of bread (regular bread: 35 g; protein enriched bread: 27 g),

^b1 portion is 150 mL,

^c1 portion is 200 mL; * Baked beans, chocolate cake, apple strudel and crackers were not consumed on the assessment days of the intervention period, and are therefore not listed in this table.

Facility did not affect intake and the data were therefore analyzed as one group. Mean protein intake increased by 11.8 grams per day (P = 0.003). This is equal to 0.18 g/kg/d (P = 0.002). Energy intake did not significantly change during the intervention period, nor did fat or carbohydrate intake (Table 4).

Table 4.

Dietary intake before and at the end of the intervention and the difference (n=22)

	Before	Intervention	Difference	P-value
Energy (kJ/d)	6856 ±1878	7139 ±1623	283	0.336
Energy (kcal/d)	1635 ± 451	1706 ± 389	71	0.314
Protein (g/d)	64.5 ± 17.7	76.3 ± 18.9	11.8	0.003$
Protein (g/kg/d)*	0.89 ± 0.20	1.06 ± 0.20	0.16	0.003$
Protein (g/kg/d)|	0.96 ± 0.19	1.14 ± 0.20	0.18	0.002$
Fat (g/d)	69.4 ± 31.7	75.7 ± 26.8	6.3	0.189
Carbohydrates (g/d)	170.5 ± 34.9	158.6 ± 37.3	-11.9	0.155

^*Unadjusted body weight was used; f Body weight is adjusted when BMI > 27 kg/m2; $ P < 0.05 indicates significance

Figure 1 shows the individual protein intakes in g per kg adjusted body weight per day of the participants before and at the end of the intervention period. The horizontal lines represent the current recommendation of 0.8 and the proposed new recommendation of 1.2 g/kg/d. Before the intervention, 19 subjects reached an intake of 0.8 g/kg/d, but only 4 reached an intake of 1.2 g/kg/d. At the end of the intervention, all subjects reached a protein intake of 0.8 g/kg/d and 9 subjects reached an intake of 1.2 g/kg/d. Only 4 subjects had lower protein intakes after the intervention than before. Using actual body weight resulted in slightly lower protein intakes per kg body weight and fewer subjects reaching the recommendations: before the intervention, 15 subjects reached an intake of 0.8 g/kg/d, whilst 2 reached an intake of 1.2 g/kg/d. At the end of the intervention, 21 subjects reached an intake of 0.8 g/kg/d, and 5 subjects reached an intake of 1.2 g/kg/d.

Figure 1.

Column scatter of individual protein intake (g/kg/d; based on adjusted body weight if BMI > 27kg/m2, this is common practice in the Netherlands) before and at the end of the intervention period

Finally, we calculated protein intake during six different meal moments: breakfast, during the morning, lunch, during the afternoon, dinner, and during the evening (Figure 2). As expected, dinner provided the major part of proteins: 29.1 g before and 28.6 g at the end of the intervention period. Only two meal moments significantly increased in protein content: breakfast increased from 12.2 to 16.0 g (P = 0.010) and the snack moment during the evening increased from 2.6 to 4.8 g (P = 0.020). Only dinner provided more than 25 g protein but this was not influenced by the intervention products.

Figure 2.

Protein intake per meal moment in g (mean and SEM are shown, *indicates a P <0.05)

Discussion

The results of this 10-day intervention study indicate that the dietary protein intake of institutionalized elderly can be increased by 11.8 grams per day with just a few protein enriched products. At the end of the intervention all elderly reached a protein intake of 0.8 g/kg/d and more elderly met the newly proposed recommendation of 1.2 g/kg/d than before (9 vs 4). The two meal moments that increased most in protein content were breakfast (+3.7 grams) and the evening snack moment (+2.2 grams). We found no evidence for a more satiating effect of the protein enriched foods, because protein intake from their regular diet remained stable.

The increased protein intake we found is in line with other studies, however ours is one of the few studies that uses protein enriched products instead of Oral Nutritional Supplements (ONS). A recent study of Stelten et al. used protein enriched bread and drinking yoghurt to increase protein intake in acute hospitalized elderly. The mean protein intake in the intervention group was 1.1 g/kg/d (75.0 g/d) compared to 0.9 g/kg/d (58.4 g/d) in the control group (25). The same protein enriched bread and drinking yoghurt were tested in older adults in a rehabilitation center for three consecutive weeks. The intervention group had a protein intake of 1.6 g/kg/d (115.3 g/d) while the control group's intake was 1.1 g/kg/d (72.5 g/d). The combination of protein enriched bread and drinking yoghurt was very effective in reaching an intake of 25 g protein during breakfast and lunch which is suggested to be beneficial for conserving muscle mass in older adults (26). Our study, however, reflects daily practice in Dutch care centers. Participants were not obliged to consume only the intervention products but were free to choose from the extensive menu options in both care centers. In addition, we offered a larger variety on product types and flavors then the aforementioned studies. Adding protein to a normal diet was also done in a study of Iuliano et al. This study included extra portions of dairy in the meals and snacks, and a difference of 25 g protein per day was reached (27). Studies that used ONS as a nutritional strategy, found inconsistent results. Neelemaat et al. found an increase of 11 g of protein per day after supplying two portions of ONS for 3 months after hospital discharge (15). Another study that used a nutrient enriched drink found an increase of 6.3 g/d with two portions of this enriched drink provided (28). In the current study, we found that using protein enriched familiar products had similar or larger effects than these studies using oral nutritional supplements. The increase in protein intake we found is comparable to most standard portions of ONS.

One of the strengths of this study is that the protein enriched products replaced products that older adults already use. This means they do not have to consume an extra serving of ONS or an extra serving of dairy or meat, as they had to do in other studies (15, 27, 28). As such, implementing this new diet was relatively easy. Also, using two different types of facilities was a strength because this provided insights into the role of free choice by the residents and the role of the awareness and actions of the personnel. This provided a realistic view of problems that could be encountered during the implementation of our protein enriched foods. It appeared that personnel was sometimes unaware of who was currently in the study and had to be offered the intervention products. The researchers observed during the hot meal that the soups, juices, mashed potatoes, and veal were only consumed when offered actively but were well accepted. This means that the awareness, attitude and actions of the personnel may greatly affect the protein intake of institutionalized elderly.

Regarding the elderly themselves, the screening for undernutrition showed that in the care home, only one participant was undernourished and two were at risk of undernutrition. In the rehabilitation center five participants were undernourished and two were at risk of undernutrition. Considering this, we think protein enriched products are especially important for rehabilitating elderly. It is known that older patients need more protein during recovery (8, 12). However, the rehabilitating elderly seemed to need more motivation or guidelines to actually consume the intervention products. This might be explained by their temporary stay while the elderly in the care home were in there own home environment and felt more at ease using products whenever they wanted. Because of the small number of participants, we should be careful when extrapolating our results to all institutionalized elderly. The small number of participants may have been the most enthusiastic, and health conscious residents. To avoid selection bias, we wanted to give all residents the intervention diet, but this raised an ethical discussion about the free choice of residents to participate in such a study. Furthermore, it can be argued that our cognitive adequate subjects do not 100 percent reflect the usual care home residents. From all 88 residents, 21 (24%) were excluded due to cognitive impairment, which is quite a large proportion. However, including cognitive impaired elderly was not possible due to the dietary assessment method which depends on memory. Therefore, the effectiveness of these products should be tested in a larger group. We are aware that the used method for dietary assessment might not be ideal. We needed an assessment method that would measure actual intake without depending on a participant's memory and cognitive abilities too much and without interfering with their food choices. For these reasons, a self-recorded food diary was not an option nor was a 24hr recall. We discussed this with several dietary assessment experts, and came to the conclusion that multiple visits per day on which a research assistant recorded food intake would be the best method. With this method the elderly just had to recall their last meal and in between snack moments. We visited the participants multiple times a day, which may have influenced the eating behavior of the participants. However, this bias does not play a role because we looked at the difference in intake between the two periods in the same subjects.

During the 10 day intervention period, subjects gave their opinion about the intervention products. Particularly the snacks, including sweet and savory pastry and baked beans, were not accepted well, and it became clear that better tasting alternatives need to be developed in the future. The change from normal bread to protein enriched bread was easily made and well accepted. We also saw an increase in juice consumption, this might be explained by the fact that the juices were freely available in the participants' refrigerators in the intervention period only. Furthermore, a larger variety of flavors and types of products can increase the liking and consumption of the products since taste and texture preferences are different per individual (29). This might also contribute to compliance to these protein enriched products on the long term.

Conclusion

In conclusion, including familiar protein enriched foods and drinks in the menu helped to meet protein recommendations in institutionalized elderly. Protein enriched familiar food products seem to be a good alternative to nutritional supplements for institutionalized elderly to reach their protein requirements.

Conflicts of interest

All three authors are employees of Wageningen University. Wageningen University received a grant from a governmental sponsor: GO EFRO.

Sponsor's Role

This research was part of the Cater with Care® project and supported by a grant from the GO EFRO 2007-2013 program (grant number 2011-016066-251111). The sponsor had no further role in the study design, data collection, analysis and interpretation of the data, writing the article, nor in the decision to submit the article for publication.

Ethical declaration:

We declare that this study complies with the current laws of the Netherlands.