This scoping review was performed in accordance with the PRISMA-ScR Checklist and the framework for scoping reviews proposed by Arksey and O’Malley [23, 24]. Even though an internal review protocol was elaborated this was not registered in any open platform. The main research question was: What is known, from literature published from 2000 onwards, about ecological determinants of eating behavior in older adults that are important to consider for developing nutrition communication strategies?

The databases used for the search were Google Scholar, Medline (via ProQuest) and PubMed. The search expression used was: (elderly OR senior OR older OR age) AND (nutrition OR food OR eating OR meals OR feeding OR “eating behavior”) AND (communication OR education OR “health promotion” OR program) AND (determinants OR factors OR influence) NOT (child OR teenager OR adolescent OR student OR young). These terms were searched exclusively in the title of publications in Google Scholar, in the title and abstract in PubMed and anywhere except the full text in Medline. The ideal approach was to search in all databases “anywhere except full text”, however this was not possible. When possible, filters were applied so that only studies pertaining to humans, older than 65 years old were obtained. The age filter was added to reduce the number of hits obtained in the search to a feasible amount to analyze. The date of the last search was the 10th of June of 2021.

Firstly, duplicate articles were removed; afterwards the title and abstracts were examined, and articles selected according to the criteria for inclusion (Table 1) and exclusion (Table 2). The full text was read for the remaining articles to assess whether these fitted the criteria. If the full text was not available, the articles were excluded. Intervention studies were excluded because the goal was to obtain an overview of determinants of eating behavior in older adults’-built environment and not yet to explore how interventions targeted these. Selected publications were summarized according to the author(s), year of publication, origin, purpose, study characteristics, outcome measurements and results.

The a-priori analysis consisted of qualitatively extracting the ecological determinants of eating behavior mentioned in the literature by reading the publications. These determinants were grouped in categories and assigned to the levels of the Ecological Framework proposed by Story et al. [12]. Categories mentioned repeatedly in literature were given bigger font sizes than the ones mentioned less often.

The software VosViewer was used for the a-posteriori analysis [25]. The software identified frequent terms among literature, these were considered to be the ones present in the title and abstract of the publications at least five times. As for the relevance score, no cut-off points were defined, the highest relevance scores obtained after the analysis were reported [25]. To analyze the evolution of research on this topic, publications were divided in four chronological periods: 2000–2004, 2005–2009, 2010–2014 and 2015–2020.

When analyzing studies published between 2000 and 2004, it becomes apparent that publications focused mainly on individual determinants [26, 27]. Terms including “attitude”, “intention” and “behavioral control” were found to influence food choice among elderly individuals [27–30] (Figure 3A).

During this period, “nutritional health” was another term found in the bibliometric mapping. This was often explored as a result of individual characteristics as sex, (oral) health and ethnicity [28, 31–36]. Vitolins et al. [33] reported that African American men aged over 70 years consumed the lowest amounts of fruits and vegetables. Some studies published during this period also reported that education, marital status, social interaction and geographical differences were correlated with diet quality among older adults [26, 31, 37–41]. In a study from van Rossum et al. [38], lower educated individuals had a lower intake of fiber and a higher intake of saturated fat and cholesterol. “Resistance to change” was also referred to as an important determinant of food choice in elderly [27]. Studies further suggest that earlier life experiences may shape present food preferences. A study from Sindler et al. [29] investigating the effects in dietary habits of survivors of extreme life events like the Holocaust, found that this experience still shaped the eating behaviors at old age. Subjects reported having reserve foods at home and to avoid food waste [29]. Regarding geographical differences, Haveman-Nies et al. [37] reported that Southern European elderly had higher diet quality than Northern Europeans.

In the following period (2005–2009), a more holistic view of determinants of eating behavior started to evolve, considering not only the individual but their built “food environment”, particularly community-related aspects. In this period, demographics as sex, ethnicity, education, income, living arrangements and geographic location continued to be reported as important determinants of quality of diet in elderly [16, 42–47]. However, these were often correlated with other collective determinants.

“Community” was a frequent term found in the publications between 2005 and 2009. It was reported that traditional foods typical of a community represent their identity and it is a complex task to replace these with healthier versions [48]. Payette and Shatenstein [16] reported that determinants of healthy eating were not only individual but also collective. Social interaction and support seem to positively influence older adults’ quality of diet, such as the intake of fruit and vegetables [16, 26]. The support of family, friends and healthcare professionals was mentioned as very important by elderly women at nutritional risk [49]. A study based in the Health Promotion model, reported that not only self-efficacy but also interpersonal support were associated with healthy eating in elderly people [50]. Underweight older adults stated that social contact during meals was crucial to maintaining healthy dietary patterns, loneliness was often associated with feelings of depression that led to lower appetite and the consumption of less balanced meals [51].

“Depression” was another term found among the literature. Payne et al. [52] compared the nutrient intake between a depressed group of elderly and a control group, the depressed group had higher intakes of alcohol and cholesterol when comparing with the control [52]. Furthermore, being positive about the aging process and perceiving fewer barriers to follow a healthy diet seemed to positively influence the consumption of healthy foods [53]. A study performed by Giltay et al. [53] reported that “dispositional optimist” in older men increased the likelihood of the consumption of healthy foods such as fruit, vegetables and whole-grain bread [53]. It has also been suggested that having food related goals as “cooking for others” were significant predictors of a varied diet [44].

“Teeth” and “functional dentition” were also frequent terms reported in the visual mapping from VosViewer. It has been shown that having a functional dentition significantly improves the consumption of fruits and vegetables in elderly males [54].

Among publications between 2010 and 2014, “oral health problem” and “dry mouth” were oral health related terms that kept on appearing at the bibliometric mapping, these lead to problems as chewing difficulty and xerostomia [55–57]. It is likely that denture wearers and/or elderly with a low number of teeth either avoid foods that would create discomfort or use food modification techniques [55, 56, 58]. Quandt et al. [56] found a significant correlation between suffering from dry mouth and having a lower consumption of whole-grain foods. It has been proposed that taste and smell impairments that accompany the process of aging are related to sensory perception of food [59]. This in turn has been suggested to affect appetite [60, 61].

There were other recurring topics in the publications between 2010 and 2014. Attitudes and beliefs towards food were, once again, shown to influence eating behavior [62–64]. Individual motivations to eating healthy were described by Dijkstra et al. [65]. Elderly reported “Health”, “Feeling fit” and “Body weight” as the most important determinants of food choice [65]. Sex, socio-economic status, living arrangements, education, ethnicity and place of residence were also described as influencing eating behavior in elderly [59–61, 65–71]. It was reported that older adults with lower education level and income adhered less to vegetables and fish/fruit government guidelines respectively [71].

“Information” was another term that appeared between 2010 and 2014. In a study performed with elderly from several European countries it was reported that even though older adults are aware of information related to nutrition recommendations and open to follow them, they draw their own conclusions from these and try to make it fit their lifestyle, habits and culture [72]. Traditional and familiar foods, often resulting from family habits and other life-course experiences, have been shown to be the preferred by the majority of older adults [72]. Several studies reported that dietary patterns of older adults are bound to their cultural and ethnic background [61]. For instance, older adults living in Mediterranean countries typically have a higher consumption of fruits and vegetables when comparing with older adults from the United States [61]. Studies further suggest that open and curious elderly are more likely to adopt novel dietary habits [73]. Older adults report that advice and provision of information from healthcare professionals have the potential to highly influence their (functional) food consumption [74]. It seems that there is an opportunity for healthcare professionals to provide nutrition information to older adults with the potential to modulate their food choices, yet this needs to be tailored to the characteristics of each target group [74]. Furthermore, it is important to understand the process of interpretation of nutrition information of different sub-groups of elderly to communicate with them in a successful way [72].

The concept denominated “Food involvement” was another term reported in the bibliometric mapping. Sommers et al. [75] investigated this concept that is characterized as “the level of importance of food in a person’s life” [75]. Pleasure was found to be positively correlated with “food involvement” [75]. Additionally, studies suggest that higher perceived stress is positively related with unhealthy eating habits such as high intakes of saturated fats and sugars, further highlighting the importance to emphasize pleasurable aspects of eating to motivate older adults to follow healthy diets [75, 76].

When exploring the last cluster of publications, “dental status” was once again a frequent term among literature. There was a focus on micronutrient intake that was significantly lower among denture wearers without any natural dentition [77]. “Education” and other socio-demographic factors were, once more, recurring themes. Several studies, suggest that commonly, highly educated older adults tend to follow or adopt healthier dietary patterns [13, 78–83].

“Motive” was another term frequently present in the most recent literature. In a study performed by Rempe et al. [84], the so called “Eating Motivation Survey” was applied to determine these motivations in older adults [84]. According to the results of this study, “Health”, “Natural concerns” and “Weight control” are among the greatest food choice influences [84]. In a study by Bardach et al. [85], older adults were asked which factors would be important for them to change their nutrition and physical activity behavior. The study revealed three main individual factors: “perceptions of old age, personal motivation and perceived confidence in ability to make effective changes” [85]. It has been discussed that it is likely that older adults with health impairments that experienced the benefits of improving health behaviors are more motivated to follow a healthy diet than the ones that do not feel that an improvement is possible [85]. Employing the ‘Extended Parallel Process Model’ it was pointed out that if individuals do not believe that they can respond to a threat (e.g., health decline due to older age) they will come to terms with it instead of trying to aim for a change [85].

“Social relationship” was also among the frequent terms in the literature mapping. Rugel and Carpiano [86] hypothesize that emotional and informational support may enhance women’s motivation to eat in a healthy way by increasing their perception of self-efficacy and sense of purpose. A lack of social relationships, particularly when associated with low economic resources appeared to influence eating behavior in older adults, especially men, in a study performed in the United Kingdom [87]. A group of older adults living in a rural area reported “community support” to be of major importance in their food choices [88]. In fact, family influence was found to positively impact the consumption of fruit and vegetables among a group of older adults from the United States [89]. Participants were familiar with food sharing and trading with friends and family; peers were often relied upon as source of assistance [88, 90].

“Guideline” was another term appearing in the bibliometric mapping of publications between 2015 and 2020. It seems that participants from lower socio-economic status perceive more barriers to follow nutritional guidelines [80]. The high price of healthy foods is often referred to as a barrier for their consumption [80, 82]. Besides, sensory appeal is mentioned in literature as important for food choice [91]. It seems that disliking healthy foods such as fruits and vegetables constitutes a barrier for their consumption by older adults, particularly, the ones with lower incomes [80, 91].

“Childhood” was another frequent term found in literature published in this period. Studies suggest that in older men, an important factor that seems to influence food choice are childhood experiences [92]. This agrees with the findings reported above regarding the influence of life-course experiences on the eating behavior of the elderly. Traditional and familiar foods, often resulting from family habits and other life-course experiences, have been shown to be the preferred foods for the majority of older adults [88, 90]. Despite this, in recent years, the food industry has started the development of novel foods adapted to the needs of older adults e.g., being of softer consistencies and fortified; however, these kind of products may not be well accepted due to the stigmatization and even difficulty accepting new foods—food neophobia [93].

A physical determinant that affects the characteristics of food environments and food choices of older adults is the difference between rural and urban settings [94]. Epidemiological studies suggest that in rural settings older adults often suffer from limited infrastructures which hinders their access to healthy food [59, 61]. Studies also suggest that transportation reliability is an important factor for accessing food/food related programs [61, 63]. It is frequently mentioned by the elderly that cheaper food shops are not accessible to them, particularly for the ones that have impaired mobility [59, 61]. The characteristics of the place of residence, including the type of foods available, cost and proximity or availability of transport to food shops, seem to influence the food consumption of older adults [59, 63, 67, 94]. Furthermore, the environment where older adults prepare and eat their meals is also of relevance [95]. A study performed by Provencher et al. [95], reported that the elderly scored better at heating soup and cutting fruit at their own homes when compared to a kitchen in a clinic. Regarding community initiatives such as meal provision and government assistance, it was reported that low participation may result from lack of knowledge and stigma around these initiatives [88, 90, 94]. The fact that meals may not be tasteful or culturally sensitive has also been reported to lead to a lower utilization of these services [63].

Several studies suggest that communication strategies need to be culturally tailored [82, 89]. A study performed by Howell [89] reported that the socio-cultural environment has a high influence on personal beliefs and attitudes towards diet. Moreover, studies suggest that the socio-cultural environment may widely vary between countries [46, 82, 89]. However, different neighborhoods may also need adapted communication strategies. A cross-sectional study from the United Kingdom investigated the influence on food choice of measures as food retail diversity, transport provision, marketing of unhealthy food and physical infrastructures that may affect mobility for older people [96]. In this study it was reported that “area (income) deprivation” is associated with the intake of fruit and vegetables among elderly [96].

Studies suggest that nutrition messages should be clear, short and targeted to the needs of specific target groups [59]. It further has been suggested that public policy interventions meant to improve the food choices of older adults should not only communicate the importance of maintaining a healthy diet but be actionable, explaining which actions can be taken by the elderly [97]. “Standalone” communication about healthy eating have been reported to be less effective than desired, but rather communication employing practical actions (e.g., nutrition workshops) may be more effective [59, 82]. Furthermore, it has been shown that interventions should focus on targeting older adults’ distrust of information sources such as food labels and media channels [74]. Additionally, it has been proposed that it is important to include the elderly in nutrition policy discussions in a citizen science approach, allowing them to share their perspectives [98].

• Adapt for sex differences as older women and men typically eat differently and may even have differences in socioeconomic status.

Research about ecological determinants of eating behavior in older adults has evolved over time, leading to a holistic view of behavior including all ecological levels. As summarized in this review, studies have reported that within this age group there are differences regarding culture, living environment and socio-demographic characteristics. These have been proposed to shape older adults’ food choice. While various theories have been developed to explain eating behavior for different age groups, for elderly, tailored measures to communicate healthy nutrition are still limited. It is important to explore the perspective of elderly towards their own behaviors and consider how to target external barriers for healthy eating.

We live in a constantly changing society; hence it is important to understand how to use the available communication resources to improve the health of this growing segment in society.

Research about eating behavior in the elderly, and ecological determinants that influence it, is crucial to achieve health and quality of life in this group. As the food environment evolve throughout the years, future research should stay updated about how various periods and events affect older adults. This will help to gain insights into the best strategies to communicate about nutrition with older adults with diverse backgrounds and modulate their eating behaviors towards healthier ones if needed.

As data about the effects of the pandemic on the eating behavior of the elderly is still limited, this review only focused on the period before that. Furthermore, only studies performed in community-dwelling older adults were included in this review. This choice was made as targeting institutionalized elderly would require investigating different stakeholders and determinants. The age range of the elderly in the studies included in this review is quite variable, yet it was decided to consider each study’s definition of older adult, as the objective of this scoping review was to gather information as broad as possible. Lastly, although an internal review protocol was elaborated, this was not registered in any open platform.

It is challenging to deliver an understandable “picture” of the determinants of eating behavior in the elderly. Here, both a-priori and a-posteriori analyses were performed with the support of graphical representations to explore determinants of eating behavior in a holistic way, thereby creating a basis for the development of nutrition communication strategies focusing on this group. Based on an in-depth analysis of the literature, this review provides advice for the development of nutrition communication strategies for the elderly. It is important to acknowledge that the topic of this review is dynamic, hence it is necessary to capture past influences in eating behavior to understand the present and make future predictions. This was done by splitting the literature into periods and analyzing patterns emerging in those.