INTRODUCTION
Malnutrition significantly impacts sarcopenia, quality of life, and mortality,1-3) with appetite loss identified as a primary contributor to undernutrition.4) In the context of community-dwelling older adults, appetite loss correlates with diminished muscle mass5) and frailty onset,6) underscoring its detrimental effects on health. Numerous factors, including physiological components such as peripheral hormones, sensory functions, central brain regulation, and impaired physical function contribute to appetite loss.7) Additionally, psychological aspects such as depression and social factors such as decreased social networks have been linked to appetite loss.
The effective management of decreased appetite requires a multifaceted approach,8) considering both physiological and non-physiological factors during assessment and intervention. Addressing social isolation and reduced social networks is imperative for combating appetite loss.9) Social networks, encompassing the functional aspects of social ties and objective characteristics like size, frequency, and density, serve as indicators of human relationships.10,11) With advancing age, individuals’ social networks often diminish,11,12) raising significant concerns regarding appetite13) and nutritional risk.14) Furthermore, social networks are closely intertwined with depression15)—a condition affecting neural circuits and hormone regulation and commonly associated with decreased appetite.16,17) In Japan, depression affects approximately 18.5% of community-dwelling older individuals,18) and it often coincides with reduced appetite. Additionally, social networks and depression have each been associated with appetite,19,20) and may exert direct and interrelated influences on appetite loss. That is, since social networks are not only associated with appetite but also serve as a factor in depressive symptoms, improving social networks may potentially enhance appetite by alleviating depressive symptoms. However, it is not clear how social networks and depression relate to each other and to appetite. Identifying this mediating effect has the potential to develop effective intervention strategies to improve appetite and, ultimately, the nutritional status of older adults. A previous study primarily explored the direct association between family and friend networks and appetite but failed to consider the indirect effects via depression.13) Given the multifactorial nature of appetite loss, a comprehensive examination that integrates factors such as social networks and depression is crucial for designing effective interventions.
In this study, we focused exclusively on social and psychological factors—specifically, social networks and depression—to address gaps in previous research and elucidate the relationship among appetite-related factors using structural equation modeling. Our primary objective was to delineate the direct effects of family and friend networks on appetite, and their indirect effects through depression. To achieve these objectives, we conducted additional studies using existing datasets. The findings of this study provide foundational insights for developing strategies to enhance appetite from both psychological and social standpoints.
MATERIALS AND METHODS
Materials
Research design and participants
This study was conducted as part of the same research project as our previous stud.13) By incorporating additional data collected from our previous research project,13) this study adopts an exploratory cross-sectional design to investigate the direct effects between family and friend networks and appetite, alongside their indirect effects through depression. The participants comprised community-dwelling older adults who attended health-checkup events held in Kasugai City, Aichi Prefecture, and Nara City, Nara Prefecture, Japan, spanning August 2019 to August 2024. A total of 229 participants—88 from Kasugai City and 141 from Nara City—were recruited, with information disseminated through community centers.
Exclusion criteria were applied as follows: (1) individuals under 65 years of age; (2) those unable to independently perform basic activities of daily living; (3) individuals requiring support or care under Japanese public long-term care insurance; (4) individuals with cognitive decline that impeded their comprehension of the study details or questionnaires; (5) individuals with significant cardiovascular, cerebrovascular, respiratory, musculoskeletal diseases, or other serious health conditions such as cancer, visual or auditory impairments, or progressive illnesses; and (6) individuals with missing data.
The initial group of 229 participants was reduced to 193 after the following exclusions: two individuals aged below 65 years, 28 people requiring support or care under Japanese public long-term care insurance, and six participants with missing data (women 78.2%; mean age 77.1±5.3 years). The reporting of the study’s results adheres to the Strengthening the Reporting of Observational Studies in Epidemiology statement.21)
Measurement Variables
Appetite
Appetite was assessed using the Japanese version of the Simplified Nutritional Appetite Questionnaire (SNAQ-J).22,23) The SNAQ-J consists of four items pertaining to appetite, each rated on a five-point Likert scale as follows:
- My appetite is (1=very poor, 2=poor, 3=average, 4=good, 5=very good);
- When I eat (1=I feel full after eating only a few mouthfuls, 2=I feel full after eating about one-third of my meal, 3=I feel full after eating half my meal, 4=I feel full after eating most of the meal, 5=I hardly ever feel full);
- I feel hungry (1=rarely, 2=occasionally, 3=some of the time, 4=most of the time, 5=all the time); and
- Food tastes (1=very bad, 2=bad, 3=average, 4=good, 5=very good).
The total SNAQ-J score ranges from 4 to 20 points, with lower scores indicating diminished appetite. This assessment tool has demonstrated its reliability and validity as an appetite assessment index.22)
Social networks
Social networks were assessed using the Lubben Social Network Scale-6 (LSNS-6)—a shortened version of the Lubben Social Network Scale.11,24) This scale comprises three items, each related to family and friend networks. The participants indicated the number of family members and friends they interacted with, those with whom they felt comfortable conversing, and individuals they considered close. The scores for each network ranged from “none” (0 points) to “nine or more” (5 points), with intervening scores of “one” (1 point), “two” (2 points), “three to four” (3 points), and “five to eight” (4 points); all scores added up to a possible total score of 30 points. A lower score indicated a smaller social network.
Depression
Depression was evaluated using the Geriatric Depression Scale 15 (GDS-15).25) This scale consists of 15 items, and participants responded to each item using a two-choice “yes” or “no” format. Each question was scored either 0 or 1, yielding a total score ranging from 0 to 15. Higher scores on this scale suggested a greater severity of depression.
Other assessment items
In addition to age, sex, and living arrangements, this study investigated the participants’ medical histories, including chronic conditions such as hypertension, heart disease, and diabetes mellitus. Body mass index, calculated as weight (kg) divided by the square of height (m), was determined based on the participants’ height and weight measurements.
Statistical Analysis
Descriptive statistics were computed to characterize the participants, with data expressed as mean±standard deviation for continuous variables and counts and percentages for categorical variables. Structural equation modeling was employed to investigate the relationship between social networks (family and friend networks), depression, and appetite among community-dwelling older adults. Given the study’s focus on the influence of social networks and depression on appetite, the observed variables included the LSNS-6 Family subscale, LSNS-6 Friend subscale, GDS-15, and SNAQ-J scores. The observed and error variables were represented by square boxes and circles, respectively. Error variables account for the variance attributed to unmeasured variables, or the factors excluded from the model. For the analysis, a hypothesis model was constructed based on clinical observations (Fig. 1).
Path coefficients were employed to illustrate the relationships between each variable. Subsequently, the hypothesized model was refined by eliminating paths with low coefficients and those lacking statistically significant associations, resulting in the final model. To conduct structural equation modeling, we ensured that the observed variables met specific criteria: skewness and kurtosis less than 2.0 and 7.0, respectively.26) Model fit was assessed using the χ2 test, goodness of fit index (GFI), adjusted goodness of fit index (AGFI), root mean squared error of approximation (RMSEA), and R2. A model was considered to have a good fit if GFI >0.95, AGFI >0.90, and RMSEA <0.05; and an acceptable fit if GFI >0.90, AGFI >0.85, and RMSEA <0.08.27) Statistical analyses were performed using SPSS Statistics 27.0 (IBM, Tokyo, Japan) and Amos 19.0 (IBM, Tokyo, Japan), with the significance level set at 5%. The target sample size for this study was estimated based on a priori power analysis (effect size=0.1, power=0.8, significance level=0.05), which indicated that a minimum of 125 participants was required.
Ethics Approval
In accordance with the Declaration of Helsinki, the objectives and details of the study, and the right to withdraw research cooperation, were explained to the participants both orally and in writing prior to the study. The research was conducted with approval from the ethical review committees of two institutions: Tokai Memorial Hospital (Approval No. 2019-003) and Naragakuen University (Approval No. 4-H026). All procedures strictly adhered to the principles outlined in the Declaration of Helsinki.
RESULTS
Table 1 presents the demographic characteristics of the participants and basic statistics for each measurement item. All four observed variables met the criteria for conducting covariance structure analysis (skewness <2.0 and kurtosis <7.0).26)
As there was no significant relationship between the direct effect of the friend network on appetite and the effect of the family network on depression, this path was excluded, and a revised model was developed. This revised model underwent testing using structural equation modeling. The goodness of fit index for the modified hypothesis model (degrees of freedom=1, χ2=0.001, p=0.975) was GFI=1.000, AGFI=1.000, RMSEA=0.000, and R2=0.22, indicating a good fit (Fig. 2). The direct effect of the family network on appetite was significant (p=0.006), with a path coefficient of 0.18. The path coefficient from the family network to depression was -0.16 (p=0.037) and from the friend network to depression was -0.25 (p=0.001). Family and friend networks had significant indirect effects on appetite via depression, with path coefficients of 0.0608 (-0.16×-0.38) and 0.095 (-0.25×-0.38), respectively (Fig. 2).
DISCUSSION
Family networks were directly associated with appetite, and indirectly associated with appetite via depression. By contrast, friend networks were not directly associated with appetite but were indirectly associated with appetite through depression. Both social networks and depression have been previously reported to be associated with appetite loss.7) However, the interrelationships between social networks, depression, and appetite have not been thoroughly understood until now. In this study, structural equation modeling was employed to elucidate the connections between social networks, depression, and appetite among community-dwelling older adults. Additionally, we differentiated social networks into family and friend networks, examining their respective relationships with depression.
We found that family networks had a direct association with appetite. Previous research has suggested that sharing meals with others can increase appetite,28) and reduced social networks have been linked to irregularities and fluctuations in food intake, as well as early satiety.29) Therefore, our hypothesis posits that networks involving family members, with whom individuals frequently share meals, directly influence their appetite. Additionally, familial relationships have been shown to impact health-promoting behaviors,30) suggesting a potential parallel influence on appetite. By contrast, the friend network had no direct effect on appetite. This could be attributed to less frequent opportunities to share meals with friends than with family members.
However, depression—a factor associated with decreased appetite—has been reported to be linked with social networks.15) We hypothesized that both family and friend networks would influence appetite through depression. While social networks tend to shrink with age, family networks have been shown to remain relatively stable in size from adolescence to older adulthood.12) Therefore, the older adult participants in this study are less likely to have experienced significant changes in their family networks unless they faced major life events such as the death of a spouse or cohabitating family member. They are also more likely to have lived in similar environments for many years. Additionally, friend networks in older adulthood can be more variable because of factors such as hospitalization, death, and changes in peer groups. For this reason, the psychological impact of the family network is likely to be limited, and the path coefficient indicating the influence on depression is expected to be greater for the friend network (path coefficient=-0.25) than for the family network (path coefficient=-0.16).
These changes in social networks may influence depression, which in turn could impact appetite. The size of a social network is also linked to the frequency of social participation.31) Studies have indicated that individuals with low social participation are more likely to face higher nutritional risks,14) and within this context, reduced social participation is likely to be associated with depression.32) Since previous research shows that increased social participation is effective for reducing depression and improving appetite,33) the findings of this study align with and support those investigations. This study underscores that family and friend networks are also associated with appetite through the mediating factor of depression, highlighting the connection between both family and friend networks (the components of social networks) and appetite. However, the mechanisms through which they are associated with appetite differ. These findings suggest that for preventing appetite loss, it is crucial to evaluate social networks separately for family and friends, and then provide tailored support for each. Specifically, individuals with limited friend networks may benefit from additional support for depression prevention alongside efforts to foster social networks.
This study has several limitations. Since the participants were Japanese, the results cannot be generalized to other countries with different cultures and lifestyles. Selection bias may have influenced the study as recruitment was conducted through a health screening program, making it challenging to ensure the inclusion of certain older adults. Therefore, future studies should aim to mitigate potential selection bias by incorporating a more diverse range of participants and employing alternative recruitment methods. Additionally, a larger sample size would be useful for enhancing the robustness of the findings. Furthermore, all the participants were self-sufficient in all aspects of daily living, which limits the ability to apply the study results to individuals requiring nursing care. While the study examined family structure, it did not assess the number of family members who regularly shared meals with the participants, which hindered a comprehensive understanding of the underlying mechanisms. This research solely focused on establishing a relationship between appetite and social networks and depression; therefore, covariates were incorporated into the analysis as error variables. Future studies should examine subject demographics and factors related to appetite, such as taste, swallowing function, and physical activity. Since this study utilized a cross-sectional design and examined the causal model using structural equation modeling, it is essential to conduct prospective, longitudinal, and intervention studies to elucidate the causal relationships between family and friend networks, depression, and appetite. Longitudinal studies would provide insights into the changes in these variables over time, while intervention studies could explore the effectiveness of interventions targeting social networks and depression for improving appetite.
In conclusion, this study investigated the direct association of family and friend networks with appetite in 193 community-dwelling older adults and explored the indirect association mediated by depression. The family network had a direct effect on appetite and an indirect effect via depression. The friend network was not directly associated with appetite but had an indirect effect through depression. Therefore, it is essential to assess social networks separately, considering both family and friends, to address appetite loss effectively. This highlights the importance of providing support for depression prevention, particularly for individuals with limited friend networks. To the best of our knowledge, this is the first study to examine the relationship between social networks and appetite by differentiating between family and friend networks. This study contributes to our understanding of how social networks and depression influence appetite—a crucial factor in the nutritional well-being of older adults—and provides valuable insights into the prevention of appetite loss.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
