Correlation Between Nutritional Decit and the Presence of Depressive Symptoms: Comparative Study Between Elderly Brazilians and Portuguese in Primary Health Care

Background: Among the countless diseases that affect the world population, malnutrition has stood out in several social and demographic contexts; among them, there has been an exponential progression in elderly people, a population group that presents a strong tendency to develop depression and malnutrition. Data from the World Health Organization show that the worldwide prevalence of depressive disorders among elderly people is in the range of 10 to 20%. Similarly, the prevalence of depression, in turn, combined with chronic diseases, accounts for 50% of premature mortality. Objective: To compare the nutritional status and the level of depressive symptoms and to evaluate the correlation between both aspects in Brazilian and Portuguese elderly people assisted in Primary Health Care (PHC). Methods: A cross-sectional, comparative study with a quantitative approach conducted among PHC elderly people in Brazil and Portugal. The sociodemographic and health questionnaire, Mini-Mental State Examination (MMSE), Geriatric Depression scale (GDS-15) and Mini-Nutritional Assessment (MNA) were used. Results: We included 160 participants (110 Brazilians and 50 Portuguese) in the study. In this sample, 81.9% of participants had a risk of malnutrition or malnutrition and 45.0% of participants had mild depressive symptoms. There was a signicant association and correlation in Brazil with the presence of a risk of malnutrition and mild symptoms of depression. Portugal also demonstrated a signicant association, however, for the chance of not presenting alterations. Conclusion: Our research showed that the risk of malnutrition or the presence of malnutrition was associated with the presence of depressive symptoms in the studied sample. In the intergroup comparison, among Brazilians, those at risk of malnutrition were more likely to have changes associated with depression, while for participants in Portugal, the odds ratio tended to show no changes, as the majority were individuals without nutritional or depressive demands. Screening and prevention measures should be emphasized in PHC to avoid the progression of symptoms in order to refer patients with medium and high complexity since this segment of the public is susceptible to malnutrition and depression.


Background
Among the countless diseases that affect the world population, malnutrition stands out in several social and demographic contexts, including when observing its exponential progression in elderly people, a population group that presents a strong tendency to develop depression, which is considered an important factor of depression, and is at risk for a worsening nutritional status (1) . Compared to healthy individuals, individuals at severe risk for malnutrition, according to the study of Leiva Badosa, Badia Tahull (2), have longer hospital stays, in addition to a 4.4-times higher risk of long-term mortality (3) .
Another challenge imposed on the world population is the pandemic of the new coronavirus 2 (SARS-CoV-2), which started in December 2019, and has had devastating socioeconomic and health effects, forcing people to con ne themselves in their homes in the rst months of its spread. This pandemic has had unprecedented important psychological and emotional impacts and immense potential to worsen depressive symptoms in the population in the coming years (4,5) .
Data from the World Health Organization (WHO) show that the worldwide prevalence of depressive disorders among elderly people reaches 10 to 20% of these individuals. Similarly, the prevalence rate of depression, in turn, combined with chronic diseases accounts for 50% of premature mortality (6) . Malnutrition in elderly people, in turn, causes major economic losses, reaching $157 billion annually in the United States of America (USA), affecting approximately three million elderly people (7) .
In Brazil, an emerging country in South America, malnutrition is strongly associated with socioeconomic factors, such as low family income, female gender and low level of education (8) ; these data are also similar when analysing depression in the same country (9) . Portugal, the country with the fourth highest number of elderly people in the European Union, presents scarce literature on nutritional assessments, especially for its geriatric population (10) ; however, mental disorders are already a real concern in relation to the entire Portuguese population (11) .
Depression is believed to be a component strongly associated with cognitive decline and malnutrition in older individuals (12,13) . With these disorders, neurodegenerative processes in speci c brain regions and of genetic factors and in ammatory processes consequently alter nutritional status and cause dementia (cerebral atrophy), which can affect regions of the brain involved in regulating appetite and eating behaviour. Atrophy of the medial portion of the temporal cortex has been associated with low body mass index (BMI), suggesting a connection between damage to the limbic system and low body weight in neurodegenerative diseases. (14) . Because diagnosis is a multifactorial and complex process, the presence of a multidisciplinary team prepared to promote welcoming and assistance to this population is essential (15) .
In this context, Primary Health Care (PHC) emerges as the focus of individual and collective promotion and prevention measures, immersed in a multidisciplinary proposal (16) . In both Brazil and Portugal, the Sistema Único de Saúde (SUS) and the Serviço Nacional de Saúde (SNS), respectively, provide the population with universal and free access to this general level of care (17)(18)(19) . PHC, which aims to promote close ties with the local community, is a privileged strategy for the development of screening, promotion and prevention measures involving diet, nutrition and mental disorders (16) . PHC was designed to prevent at least 60% of health problems, including depression (20) .
The screening, diagnosis and treatment of depression, despite the increase in its prevalence, remain de cient in clinical practice (21) . However, when identifying the need for hospitalization due to malnutrition, elderly PHC users should be referred to secondary or tertiary care, a scenario in which the level of malnutrition in elderly people was found to be twice as high compared to those treated at the hospital itself (22) . This problem, in addition to the growing incidence of depression and malnutrition in elderly PHC users, reinforces the importance of more effective measures, as elderly people constitute one of the groups most affected by the factors associated with these diseases. Elderly people are seen in health units, both in Brazil and in Portugal, because these services are among the main entry points for health in both countries. However, there are no specialist professionals, such as nutritionists and psychologists, because they are not mandatory members of teams of health professionals in basic health units (17) . To provide support, the Family Health Support Center, FHSC (23) (in Brazil), and the Shared Assistance Resource Units (24) (in Portugal) were designed to promote more effective referrals since the specialist professionals working in these centres work within communities. However, these centres operate at low capacity because they have few human resources and because they are not a mandatory in all areas covered by PHC (17) . Therefore, we start from the hypothesis that the risk of malnutrition or the presence of malnutrition in uences depressive symptoms among elderly people.
Our objective was to compare the nutritional status and the level of depressive symptoms and to evaluate the correlation between both aspects in Brazilian and Portuguese elderly people assisted in PHC. participants were asked to sign free and informed consent forms as a way of accepting participation in the study as well as authorizing the use of their information.

Design and Study Site
A cross-sectional, comparative study using a quantitative approach was carried out with elderly people assisted in Primary Health Care (PHC) in Family Health Strategy Units (FHS) in the cities of Natal and Santa Cruz, both located in the State of Rio Grande do Norte, Brazil, and in Elderly Health Care Services (EHCS) in the coverage area of Family Health Units (FHU) in Eborae, Plaicie and Salus, linked to the regional health administration of the Municipality of Évora (Council Évora), in Portugal.
The following inclusion criteria were used: 60 years of age or older in Brazil (25) and 65 years of age or older in Portugal (26) according to the legislation in force in each country; registration in or being a user of a Family Health Unit in one of the study sites (Brazil / Portugal) for at least six months from the day of data collection; and a score of at least 17 points on the Mini Mental State Examination (MMSE) questionnaire, as suggested in its validation (27) . The following exclusion criteria were applied: history of limb amputation and/or a physical inability to remain upright for anthropometric measurements; medical diagnosis of an intellectual or mental disability.

Instruments
The following instruments were used in the present study: sociodemographic and health questionnaires, including closed questions (age, sex, education, family income and marital status, in addition to questions about health, such as the presence of chronic diseases); the MMSE questionnaire (28) , which is composed of several items that assess cognitive functions and is used as a cognitive screening tool; the Geriatric Depression Scale (GDS-15) (29) , which includes 15 objective and dichotomous questions (yes/no) about how the interviewee felt in the previous week using the following cut-off points to interpret the results: 0 to 5 points (no depression), 6 to 10 points (mild depression), and 11 to 15 points (severe depression); and the Mini Nutritional Assessment (MNA), which is composed of 18 questions whose total score differentiates the following groups: MNA > 24 (adequate nutritional status), MNA from 17 to 23.5 (at risk of malnutrition), and MNA < 17 (total malnutrition) (30) Participants' weight was measured using a portable digital platform scale with a maximum capacity of 150 kg and an accuracy of 100 g. Height was measured with a 5-meter retractable steel tape measure on a wall on which the participants were standing, positioning their heads according to the Frankfurt plan and following the procedures of the Food and Nutrition Surveillance System (FNSS) (31) . The circumferences of the arm and calf were measured using an inelastic measuring tape with an accuracy of 1 mm (32) .

Data collect
Data were collected by undergraduate and graduate students in Nursing and Health Sciences with Master's and doctoral degrees, including nurses, nutritionists and physiotherapists who were previously instructed and trained, without remuneration, for the handling of anthropometric instruments and measurements. Training on the application of the instruments was carried out by nurses (for the sociodemographic questionnaire and GDS-15) and nutritionists (for MNA and anthropometry) who had previously used these questionnaires in clinical practice and/or in research. They did not participate in the study data collections. in Évora, as requested by Portuguese participants. Neither the participants nor the interviewers were blinded.

Statistical analysis
The data were tabulated with the aid of Microsoft® Excel 2016 (Microsoft Corporation, Washington, WA, USA). The Statistical Package for the Social Sciences (SPSS) software (IBM, Armonk, NY, USA) version 20.0 was used for statistical analysis of the data, and the normality of the data was assessed by the Shapiro-Wilk test, which indicated non-normality of the sample, followed by the descriptive and nonparametric Chi-Square test and Fisher's exact test for sociodemographic data and associations, respectively. In addition, Spearman's correlations and chance ratios between nutritional status and depressive symptoms were tested. The correlation levels (Spearman's Rho Test) were categorized as weak (r < 0.40), moderate (r 0.40-0.49), and strong (r > 0.50). The signi cance of all the ndings were established as ρ-value ≤ 0.05.

Results
The study had a total sample of 160 participants, including 110 Brazilians and 50 Portuguese. Among the 118 Brazilians approached, eight participants were excluded, four for not reaching the minimum score of the MMSE, two for being registered in their health unit for less than six months and two for having physical limitations that did not allow the measurement of anthropometric data. In the Portuguese group, there was no exclusion of participants. Table 1 shows the predominance of women (75.0%) and younger individuals aged between 60 and 80 years (88.2%), as well as the predominance of chronic diseases (84.4%) in both countries. In Brazil, most participants had no partner (51.8%) in contrast with the group in Portugal (44.0%). In terms of income, 54.5% of Brazilians earned 1 minimum wage or more, while 76.0% of Portuguese showed a maximum income of up to 1 minimum wage (ρ-value < 0.001). In both groups, those with 1 to 5 years of schooling predominated (ρ-value < 0.001). In Brazil, there was a higher frequency of participants with changes in both nutritional and depressive symptoms (Table 2), with a chance of changes of 17.5 (ρ-value = 0.001). In Portugal, the trend for individuals with normal nutritional status and without depressive symptoms, that is, the chance of not showing any changes, was 21.9 (ρ-value = 0.001).  Table 3 shows that in Brazil, when analysing nutritional status and depression levels, the highest frequency was found for individuals with mild depressive symptoms and who were at risk of malnutrition (ρ-value = 0.003). In Portugal, individuals with normal nutritional status and without depression predominated (ρ-value = 0.007). In general, the participants were at risk of malnutrition and mild depressive symptoms, with a signi cant difference in the analyses.  In the analysis between nutritional status and depression, the correlation (Table 4) between both aspects was found to be signi cant and weak (r = 0.36) in individuals with the presence of nutritional changes and the presence of depressive symptoms (ρ-value < 0.001). Signi cance was also found in both countries; however, in Brazil, the parameters were directly proportional, that is, the more nutritional changes found, the more the depressive symptoms became evident, although with a weak correlation (r = 0.23). In Portugal, the results showed a trend towards normal nutritional status and without depression, with a strong correlation (r = 0.51).

Discussion
Our study showed that the risk of malnutrition or the presence of malnutrition was associated with the presence of depressive symptoms in the studied sample. We also observed a predominance of participants at risk of malnutrition and mild depressive symptoms in Brazil, with a 17.4-fold chance of presenting nutritional or depressive changes and a signi cantly weak and positive correlation. In Portugal, we observed a tendency towards the absence of changes in both nutritional status and depressive symptoms, with a high chance of not having changes (21.9 times) but with a strong positive and signi cant correlation.
These ndings may be related to the pro le of greater female predominance and higher levels of education found among the Portuguese participants. A study conducted by Ghimire, Baral (33), who used a pro le similar to ours, suggested that this reality is explained by the fact that compared with men, women seek more access to health services in addition to placing a greater value on the preservation of their autonomy in daily activities and self-care, a fact also in uenced by a higher level of education.
The presence of chronic diseases in our sample indicated to be predominant in both scenarios, although there was no prominence between them, which may explain the fact that Brazilians had a considerable chance of developing nutritional and depressive changes. In this regard, it was estimated that individual who live with chronic diseases have a greater tendency to experience reduced quality of life and life expectancy due to the negative consequences exerted by them (34) .
From a nutritional point of view, other studies have inferred that patients with chronic and malnourished diseases have a signi cant impairment in quality of life in addition to being more depressed and having lower life expectations (35,36) .
Regarding monthly income, in Portugal, some individuals did not have a regular income, and no participant had an income above the minimum wage, a situation opposite to that in Brazil, where participants had a higher level of income, despite their worse nutritional and depressive assessments. It is known that the quality of food is usually associated with how much income an individual or family has, and consequently, this aspect is inversely related to the risk of malnutrition/malnutrition (12,13) . Governmental measures for the selection and inclusion of these groups in income transfer programs, especially the Social Insertion Income (SII) in Portugal, can minimize the risk of malnutrition/malnutrition among elderly people, which aims to alleviate poverty, improve living conditions and health and encourage healthy habits (34) .
However, these aspects need to be addressed by the PHC in an educational and consistent way by promoting an orientation towards healthy nutrition consistent with the reality of each individual, family and community (19) .
Other studies have shown that both low income and low education are associated with the risk of malnutrition and therefore may increase the risk of food insecurity (34,37) , which was one of our ndings as well. Alvares and Amaral (38) included 3,552 Portuguese individuals; of these, 16.5% exhibited food insecurity in association with the same variables (38) . Food insecurity is characterized as an individual who has a limited or uncertain possibility of acquiring food in a socially acceptable way or when the availability of food is compromised in relation to nutritional adequacy and safety (39) . Thus, these factors may have contributed to the highest percentage of participants with poor nutritional status found in the two groups in our study.
Research carried out in Switzerland showed a higher percentage of individuals with depression, who were more prone to risks of malnutrition and frailty (40) . This reality was indicated in the two countries we studied, but more clearly among Brazilians. Regarding this scenario, the importance of multidimensional interventions is increased to improve and/or minimize the nutritional de cits and depressive symptoms of the participants, as illustrated by successful past experiences in the Brazilian PHC scenario itself (9) . However, the coronavirus pandemic, which began in 2019, is also projected to have effects on the general population, which includes elderly people, involving greater di culty in circumventing the problem of depression in view of its substantial impact on mental health, which is generated by economic losses, deaths of family members, and other aspects affected in terms of behaviour, lifestyle and, consequently, nutritional status (4,5) .
Individuals in our study who were at risk of malnutrition had mild depressive symptoms in both countries, corroborating other ndings in Taiwan, which also signi cantly associated the risk of malnutrition with depressive symptoms, where similar scales were used to measure these aspects, however with elderly people considered fragile (39) . Our results may be associated with poor eating habits due to loss of appetite, which is aggravated by decreased food intake and lack of palatability and usually causes malnutrition with the risk of increased malnutrition when associated with depression (41) .
In addition to showing an association and correlation between nutritional status and depressive symptoms, our study found a predominance of risk of malnutrition and mild depression in Brazil in comparison with Portugal. In a study conducted in Nepal, 10% of elderly people surveyed were malnourished, which accounted for 41% of the total effect of depression on quality of life, while depression had a 6% share of the total effect of malnutrition (33) . Similarly, the Brazilians in our sample appeared to suffer the same greater in uence of malnutrition on depression than vice versa. That is, we observed a proportionally higher percentage of Brazilians with strictly nutritional demands than those who had only depressive symptoms.
Evidence indicates that good eating habits can mitigate depressive symptoms (42,43) . In the study of Jacka, O'Neil (44), 32% of patients in two Australian hospitals who experienced moderate to severe depression and underwent a three-month dietary intervention showed signi cant improvements and remission of depressive symptoms. Therefore, the initiatives and interventions carried out in PHC units, which are gateways for the population, constitute a key strategy for health measures for active aging, which makes it possible to improve nutritional status and prevent malnutrition and depressive symptoms in medium and high complexity units. Active aging stands out as a concept of physical, social and mental well-being throughout life, that is, adopting healthy dietary, mental, exercise and leisure habits (45) .
PHC, with integrated care and networks, therefore, is vital for the promotion of active and healthy aging, as well as for the resolution of basic health demands in both countries. Collective measures are essential, mainly because they increase user participation in the educational process, team involvement with the participant and work optimization (17,21,46) . With the ndings of two different scenarios, the importance of identifying changes in both nutritional status and mental health in the PHC scope is observed to improve the quality of life of the population towards healthy and active aging.
One of the limitations of this study was the sample size, especially of the group from Portugal, which suggests bias in the comparison between countries and impairment of the intergroup isonomy. In addition, the cross-sectional design of the study makes it di cult to establish causality. Thus, the inference power of our results is limited. Despite these obstacles, the methodological path we committed to was rigorously applied to seek the maximum similarity among the participants studied and, consequently, the greatest consistency of the data and its interpretations with respect to economic, social, cultural and cultural characteristics.

Conclusions
Our research showed that the risk of malnutrition or the presence of malnutrition was associated with the presence of depressive symptoms in the sample studied. In the intergroup comparison, among Brazilians, those at risk of malnutrition were more likely to have changes associated with depression, while for the participants from Portugal, the odds ratio tended to show no changes, as the majority were individuals without nutritional or depressive demands. In general, most participants were at risk of malnutrition/malnourishment, which was strongly associated with low income and low education. These ndings allowed us to indicate our research hypothesis.
Screening and detecting malnutrition and depressive symptoms should be used as measures to prevent these problems. We believe that our results will help PHC professional teams to develop promotion and prevention measures not only for malnutrition and depression but also for other disorders in addition to public policies. The importance of the presence of a multidisciplinary team of nutritionists and psychologists as direct and indirect support in PHC is also emphasized to mitigate repercussions, thereby avoiding the need for hospitalizations and worsening cases.
GAST, BASD and MAGJ contributed to the study design, data acquisition and writing of the manuscript. JMAM, SMSGSOT, NPLT and ACVC contributed to the study design, acquisition and analysis of data and preparation of the manuscript. TTXN, FRPM, FANM, GVT and EMCM contributed to the analysis, draft and critical review of the manuscript. All the authors read and approved the nal manuscript.