Impact of lifestyle habits on the health self-assessment of rural older men of different ages: evidence from the Chinese longitudinal healthy longevity survey

2.1 Sources of information

Cross-sectional data were obtained from the 2018 Chinese Longitudinal Healthy Longevity Survey (CLHLS), which is a pioneering and extensive gerontological and social science survey project in China managed by the Peking University Center for Healthy Aging and Development. As the earliest and most comprehensive nationwide survey of its kind, CLHLS covers 23 provinces, municipalities and autonomous regions, making it a critical resource for analyzing gerontological health issues. Employing a targeted random sampling approach, the CLHLS focuses primarily on individuals aged 60 years and above. With a database comprising 113,000 household interviews, the scientific rigor and representativeness of the sample selection are well-established [17]. From the 2018 CLHLS dataset, which included 15,874 interviews, we selected data on male adults aged 60 years and above residing in rural areas. After excluding samples with missing key variables or inconsistent responses, 1996 samples were retained for inclusion in our study analysis.

2.2 Research methodology

In this study, the primary outcome variable was the self-assessed health status of elderly rural men, which was measured through the survey question: “How do you feel about your own health now?” serving as a subjective indicator of the participants’ overall health perception. To improve the analysis of perceived health outcomes, we utilized a 4-point scale, aligning with previous research that suggests such a scale effectively distinguishes perceptions of health into positive and negative categories [18]. Accordingly, we coded responses of “good” and “very good” collectively as “good”, assigning them a value of 4; “fair” received a value of 3; “not good” was coded with a value of 2; and “very bad” was assigned the lowest value of 1.

In this study, the independent variables were composed of five lifestyle habits: smoking status, alcohol consumption, sleep duration, engagement in physical exercise, and frequency of social interactions. Additionally, several control variables were included to account for individual characteristics, namely age, marital status, and education level of the respondents. The classification of smoking habits was based on current smoking status. Participants who confirmed being currently smoking were categorized as engaging in the harmful habit of smoking and were assigned a value of 1; otherwise, they received a value of 0. This categorization is grounded in evidence indicating that both light (fewer than five cigarettes per day) and heavy smokers (five or more cigarettes per day) exhibit a significantly accelerated decline in lung function compared to non-smokers [19]. For alcohol consumption, the CLHLS collected data on average daily intake and the primary type of alcohol consumed. Following methodologies from prior studies [20], we calculated the average daily alcohol consumption for each respondent. Consistent with research suggesting that a daily intake of up to 10 grams of alcohol may be considered safe [21], consumption exceeding this threshold was deemed excessive drinking and assigned a value of 1; all other instances were coded as 0. Sleep duration was categorized based on health recommendations, with 6–8 hours of sleep considered normal (assigned a value of 1) and durations of less than 6 hours or more than 8 hours considered abnormal, receiving a value of 0 [22]. Regular physical activity was identified through the question, “Do you do exercises regularly at present?”. Affirmative responses indicated regular physical activity and were assigned a value of 1, while negative responses were coded as 0. Educational attainment was quantified based on the total years of education completed by the respondent, with specific benchmarks (e.g., completion of elementary school) receiving corresponding numerical values. Age groups were delineated according to World Health Organization criteria, dividing the rural elderly men into low-aged (60–74 years), middle-aged (75–89 years), and high-aged (≥90 years) categories [23]. Table 1 presents the assignment of values for each variable and descriptive statistics, offering a comprehensive overview of the variables under study.

2.3 Statistical methods

In this study, we first determined the association between each variable and the self-reported health status of elderly rural men via one-way analysis of variance (ANOVA). Subsequently, multiple regression analysis was performed to test whether the factors were still significant in the presence of multiple other factors and whether there was age heterogeneity in the strength of the effects. Finally, we clarified the changes in the significance of the effect of the factors on health self-assessments of age groups using the decomposition of Shapley value. One-way analysis of variance was performed using the Chi-squared test and Kruskal-Wallis rank sum test depending on the type of independent variable. Regression analysis was performed using multivariate ordered logistic regression since the dependent variable was a multivariate ordered variable. Shapley value decomposition is an important method for calculating the difference in the contribution value of each independent variable to the dependent variable by decomposing the R² in the framework of cooperative games, thus distributing the benefits. Furthermore, it is a reliable tool for effectively decomposing the causes of problems in social sciences [24] and can be used to demonstrate the specific variations in the strength of the influence of lifestyle habits on the health self-assessments of rural men in old age in different age groups. Data processing and analysis were performed using Stata 16.0 (StataCorp LLC, Texas, USA). A p value < 0.1 was considered statistically significant.

3.1 Basic information of the respondents

Among the 1996 elderly rural men included in this analysis, 1732 (86.8%) reported positive health self-assessments, while 264 (13.2%) reported negative health self-assessments. The mean age of participants was 82.15 ± 10.80 years. The age distribution was relatively even across the designated age groups, with 589 (29.5%) respondents aged 60–74 years, 811 (40.6%) aged 75–89 years, and 596 (29.9%) aged 90 years or older. In terms of marital status, 1186 (59.4%) of the respondents were married, and 810 (40.6%) were unmarried. The average educational attainment among the participants was 3.79 years, reflecting a generally low level of literacy in the sample.

3.2 One-way analysis of the impact of lifestyle habits on the health self-assessments of rural men in old age

The results of the one-way analysis showed that except marital status, all other lifestyle habits, including smoking status, drinking status, sleep duration, physical exercise, frequency of friend interaction, age and educational attainment, were all significantly associated with the self-assessed health status of rural elderly men (p < 0.1) (Table 2).

3.3 Regression analysis of the impact of lifestyle habits on the health self-assessments of rural men in old age and age group differences

To examine the impact of lifestyle habits on the self-reported health status of rural elderly men across different age cohorts and to assess the presence of heterogeneity when all factors are considered together, this research employed a multivariate logistic regression model for multifactorial analysis, with health self-assessment as the outcome variable. Given that marital status did not exhibit a significant relationship with health self-assessment in the preliminary analysis, it was excluded from the regression model. The results of the regression analysis indicated that maintaining normal sleep duration (β = 0.305, p < 0.01), engaging in regular physical exercise (β = 0.318, p < 0.01), and frequent interaction with friends (β = 0.162, p < 0.01) acted as protective factors for the health of elderly rural men. Conversely, excessive drinking (β = −0.550, p < 0.01) was identified as a detrimental factor to their health. Furthermore, the analysis revealed considerable variation in the impact of these factors on the health of rural elderly men across different age groups (Table 3).

3.4 Shapley value decomposition of the effect of lifestyle habits on the health self-assessments of elderly rural men in different age groups

While regression analysis reveals the direction and strength of the impact that independent variables have on health assessments, it does not fully explain how the importance of each factor changes across different age groups. To overcome this limitation, we applied the Shapley value decomposition method, which allowed us to include factors that showed significant effects in the regression analysis and assess their varying impacts on the health perceptions of elderly rural men across low, middle and high age categories. The results indicated that the impact of lifestyle habits on perceived health status differed significantly by age group: drinking status influenced the low age group by 53.27%, the middle age group by 35.40%, and the high age group by 10.63%; sleep duration affected them by 41.17%, 8.93% and 7.68%, respectively; physical exercise by 3.03%, 24.77% and 57.03%, respectively; and social interactions with friends by 2.52%, 30.90% and 24.66%, respectively, as shown in Table 4.

4.1 Lifestyle habits have a significant impact on the health self-assessments of rural elderly men

The impacts examined in this study are focused on perceived health, as the data collected is based on self-reports, making it inherently subjective. This methodology does not juxtapose the respondents’ self-assessments with any standardized objective health metrics. However, despite the potential variances between self-reported health status and clinically assessed health status, perceived health remains a crucial indicator of an individual’s health condition [25, 26]. Gerontological research substantiates that the self-assessment of health by older adults can reliably predict their actual health status, with markedly negative self-perceptions of health being associated with an increased risk of premature death [27].

The present study identified a notable influence of lifestyle habits on the self-perceived health status of rural elderly men, aligning with the findings of Ye Jinzhen [5]. It has been established that smoking and excessive alcohol consumption are significant risk factors for deteriorating health in senior years [28, 29]. However, our regression analysis indicated that smoking did not significantly impact the health self-assessments among this demographic, possibly due to the confounding effects of other variables. Additionally, it was observed that a considerable proportion of older men engage in excessive alcohol consumption [30, 31], underscoring the need for targeted interventions in this population group. Excessive alcohol consumption among rural elderly men is linked to poorer self-assessments of health, a relationship potentially rooted in the association between heavy drinking and the onset of common mental disorders, immediate alcohol-related adverse effects, and dependency on alcohol [32]. These negative outcomes of excessive alcohol use are likely to impair older adults’ perception of their health. Furthermore, the duration of night-time sleep serves as an indicator of the rest quality obtained by the elderly. Insufficient sleep can result in daytime lethargy, while oversleeping may cause prolonged oxygen deprivation and neuronal inhibition in the brain, leading to feelings of fatigue and emptiness. Thus, sleep irregularities can diminish immunity in the elderly, increasing vulnerability to various health conditions [33]. Our findings also highlight that rural elderly men engaging in regular physical activity reported better perceived health, aligning with the research of Yang Fan et al. [34]. Additionally, engaging in deliberate physical exercise not only enhances physical well-being but also alleviates psychological stress. The frequency of interactions with friends positively correlates with the self-reported health status among rural elderly men. Active social engagement not only wards off the decline in motor functions but also significantly enhances both physical and mental health, bringing about mental contentment [35]. To capitalize on these insights, township government departments should collaborate with community organizations to highlight the significance of lifestyle habits for elderly health. This collaboration could involve inviting experts to deliver comprehensive lectures on fostering positive lifestyle habits in rural areas and offering programs and services designed to meet the specific needs of rural elderly men, aligning with supportive public policies. Establishing village gymnasiums for the organization of low-intensity physical activities, such as jogging and tai chi, by village officials, as well as cultural rooms for engaging in cultural and recreational activities like calligraphy, chess, singing and dancing, could provide the elderly with the means to translate knowledge into action. This initiative aims to facilitate a shift in mindset and behavior among the elderly population [36]. Additionally, rural family members are encouraged to pay closer attention to the elderly, reinforcing their advice and encouragement to foster healthy living habits. For their part, rural elderly men should strive to limit alcohol consumption, maintain regular sleep patterns, engage in active exercise, and continue socializing to improve their health.

4.2 Effects of alcohol consumption and sleep hours on the health self-assessments of rural men in old age diminish with age

Excessive drinking significantly affects younger male seniors more than older ones, potentially due to alcohol’s effect on nerve function, leading to motor and cognitive impairments in older adults [37]. Younger seniors, who are more active in daily behaviors such as working and socializing, may more acutely feel the detrimental effects of alcohol. In contrast, middle-aged and older seniors who engage less in physical activities might not perceive the negative impacts of excessive alcohol use as strongly. Nonetheless, the reduced influence of excessive alcohol consumption on the self-rated health of older seniors does not diminish the actual health risks associated with heavy drinking, especially considering the natural decline in physical health that accompanies aging. The influence of sleep duration on perceived health diminishes with age, likely because younger rural elderly men are more engaged in physical labor [38]. Their higher level of physical activity demands quality rest to recover from the energy expended, making adequate sleep crucial for their perceived health improvement. In contrast, for middle-aged and older seniors, whose physical activity is less, the necessity for extensive rest decreases [39]. Consequently, the significance of sleep for their perceived health perception diminishes. It is important to convey the benefits of healthy behaviors to this demographic through engaging methods, such as memorable slogans like “Don’t touch alcohol, be healthy” and “Sleep well, be healthy together”, and through educational content like films and expert lectures. Township government departments could implement measures to limit unnecessary “drinking parties” in villages to prevent younger seniors from developing drinking habits in group settings that are hard to break later in life. Moreover, creating village regulations, with input from residents, to define acceptable hours for night-time activities can help manage disturbances. Proactively addressing activities that fall outside these hours is essential for minimizing rural noise pollution, thereby preserving a conducive sleep environment for the elderly and enhancing their sleep quality. Additionally, families in rural communities should play a supportive role in moderating the drinking habits and sleep patterns of the elderly, who in turn should be encouraged to actively participate in these health-promoting practices.

4.3 The impact of physical activity and friends’ interaction on the health self-assessments of elderly rural men strengthens with age

Compared with their younger counterparts, senior male elderly exhibit a significant decline in self-care and socialization abilities. This decline contributes to increased feelings of loneliness, worthlessness, and other negative emotions [40], which, if unresolved, adversely affect their perceived health. However, the potential for enhancing the quality of life through physical exercise remains significant, and the ability to maintain social connections does not necessarily decrease with diminished physical function [41]. Aging individuals often rely on physical activity and social interaction as primary methods for alleviating stress and improving well-being. Participation in group exercise further facilitates extended periods of communication among participants, effectively reducing physical strain and alleviating psychological stress through social support [42]. Thus, it is advisable to promote suitable physical activities for rural elderly men, particularly those of middle and advanced age. Organizing regular, low-intensity group exercises and social activities could substantially benefit their physical and emotional health. Specifically, rural communities can increase older men’s engagement in activities by introducing incentive mechanisms like “punch cards” and “invitations”. Through these systems, older men can accrue points by attending group activities or by inviting fellow peers, with additional rewards for including the eldest among them. These points can then be redeemed for practical items like eggs, grains, oil and seeds. It is important for family members to collaborate with the community, clarifying how these activities work and their benefits and motivating elderly men to join. They should also offer guidance on the appropriate level of physical activity and its duration. In turn, elderly men are encouraged to embrace the support from their communities and families, actively engaging in and exploring new activities.

Our study has some limitations. The present study faces several limitations. Firstly, the evaluation of participants’ health status was based solely on self-reported questionnaires, lacking clinical corroboration. Secondly, constrained by the questionnaire’s design, our analysis was limited to assessing the impact of current smoking and excessive alcohol intake on self-reported health. The comprehensive aspects of these behaviors, such as frequency and type of consumption, and their specific effects on the health of rural male elderly, remain to be elucidated in future research endeavors. Thirdly, the employment of cross-sectional data precludes the derivation of causal inferences. The observed differences across age cohorts might not only reflect the intrinsic attributes of these groups but could also be indicative of cohort effects, arising from societal evolution and changing lifestyle patterns across generations. Consequently, the generalizability of our findings is limited, necessitating further longitudinal studies to confirm the long-term health consequences associated with detailed lifestyle habits.