Neighborhood’s locality, road types, and residents’ multimorbidity: evidence from China’s middle-aged and older adults

Despite that extensive studies have focused on the urban vs. rural disparities in multimorbidity, the vast majority of existing research did not control for neighborhood-level variations [19‐23]. For example, there are two relevant studies based on the Chinese population, the results of which suggest that urban residents have higher odds for multimorbidity [21, 24]. Both studies, however, employed between-person analyses, which did not allow them to adjust for neighborhood-level variations and any neighborhood characteristics [21, 24]. Whether the urban-rural disparities in multimorbidity remain after controlling for neighborhood-level variations is unknown.

Living in walkable neighborhoods has been recognized as a protective factor of residents’ health. For example, living in a neighborhood with access to greenness and open space may provide residents a space for walking and social coherence, and therefore lead to better mental and physical health [25, 26]. The effect of living in a walkable neighborhood on residents’ health as a reflection of non-communicable chronic disease is worthy of exploration, particularly in China, where the vast majority of individuals in China utilize in-center health services. In addition to the effect on social coherence and recreation [25, 26], living in a walkable neighborhood could play a key role in residents’ access to health care and ultimately affect the individuals’ regular and long-term chronic disease.

However, little research has investigated the association between walkable neighborhoods and multimorbidity in China, for which the adoption of walkability measures is a major issue. Prior studies have defined the extent to which the neighborhood’s built environment affects walking as walkability [27]. Walkability incorporates a variety of measures including but not limited to land use mix [28, 29], residential density [28, 29], access to recreational and green space [25], access to transportation [30, 31], and street connectivity [32]. Various indicators of walkability have been proposed, including subjective perception and objective measures that employed individual surveys or GIS-based approaches [27, 29, 33‐35]. However, it is well-acknowledged that a “one-size-fits-all” indicator of walkability does not exist. Although a prior study has modified and validated Neighborhood Environment Walkability Scale-Abbreviated (NEWS-A) among Cantonese-speaking older adults in Hongkong, the culture and city planning differ substantially between Hongkong and other cities in China [33]. The validated version should be cautiously administered in mainland China before further modification and validation. Moreover, the existing measurements that developed in urban cities may not be applied to rural China, where well-defined aspects of walkability, such as access to recreational space and supermarket, maybe inappropriately examined because of the inadequate facilities.

The present study proposed road types as a potential indicator of the neighborhood’s walkability for the following reasons. China has launched several ambitious plans to improve rural infrastructure, of which investment in rural roads dominated a considerable share [36]. Although the expansion of rural roads has linked rural residents to the outside world, where job opportunities, medical services, and schools are, the quality of roads has raised serious concerns [37, 38]. Upgraded or paved roads may decrease time costs for access to health services, particularly in less developed countries where the rainy season would render unsurfaced roads impassable [38‐40]. Given that the association between access to health care and the population’s well-being has been received [41‐43], and there is initial evidence linking road constructions with residents’ cognition [18], living with paved roads may be associated with residents’ multimorbidity, which requires substantial regular and long-term health care services [22, 44]. However, there exists limited evidence directly linking neighborhoods’ road types with multimorbidity, and the present study aims to address the gap.

Overall, this study aims to investigate the association between two neighborhood characteristics and multimorbidity among Chinese middle-aged and older adults. We offered two hypotheses for data analyses. First, urban-rural disparities in multimorbidity may not be consistent after adjusting for neighborhood-level variations (H₁). Second, living in a walkable neighborhood may be associated with fewer NCDs (H₂) since walkable neighborhoods could increase residents’ access to health care [45]. Here, we measured road quality by differentiating roads into paved roads vs. unpaved roads to reflect the neighborhood’s walkability. We defined the paved road as a more walkable road compared with the unpaved road. However, cautiousness should be taken when generalizing our findings into the literature of walkability in relation to health.

We employed a cross-sectional study. Data came from the China Health and Retirement Longitudinal Study (CHARLS) 2011 National Baseline Survey, which included 17,708 Chinese aged over 45 [46]. Individuals came from 28 provinces in mainland China and were recruited by a multistage random sampling procedure [46]. Information regarding sampling, recruitment, response rate, and procedures for data collection could be retrieved from a prior study [46]. The present study used individual-level data from the individual survey as well as neighborhood-level characteristics from the community questionnaire from CHARLS. In our definition, neighborhoods refer to villages in rural areas and communities in urban areas. As CHARLS was a community-based survey, individuals were clustered within their neighborhoods (communities or villages) by sharing the unique neighborhood identification. Excluding 4295 observations with missing data, we derived data on 13,413 subjects from 432 neighborhoods.

First, we stratified study subjects into urban and rural groups to exam the distribution of the baseline characteristics. T-test was employed for age. Mann–Whitney U tests were employed for the number of primary care institutions, educational attainment, household income, BMI, and physical activity. Chi-squared tests were employed for road types, groundwater systems, sex, marital status, and health care insurance. The analyses were used for in-sample interpretations; thereby, CHARLS survey weights were not introduced (Table 1).

When estimating the prevalence of multimorbidity and each NCD among China’s middle-aged and older adults, CHARLS complex sampling weights were used to account for selection bias (Table 2). Since there were no significant differences between results from weighted analyses and unweighted analyses, unweighted analyses were used for the following modeling analysis.

Negative binomial regression was employed to investigate variations in multimorbidity. We employed negative binomial regression rather than Poisson regression since the dependent variable’s variance was larger than its mean value. Univariate analyses were performed to examine the association between multimorbidity and each independent variable, respectively. Multivariate negative binomial regression analysis was employed with all covariates introduced. We performed this between-person analysis (Model 1 in Table 3) in order to compare our results in terms of urban-rural disparities with those from prior studies since most of them did not adjust for neighborhood-level variations [21, 24]. Clustered robust standard errors were generated in Model 2 to take individuals nested within neighborhoods into account. Furthermore, we employed the separate logistic regression models with robust standard errors to investigate the association between neighborhoods’ characteristics and the fourteen chronic diseases, respectively (Table 4). Variance inflation factors were calculated to exam collinearity among independent variables. Models’ significance was examined by Pearson chi-square.

Table 1 presents the baseline characteristics of the study sample. Overall, the present study included 13,413 individuals with 5639 urban residents and 7774 rural residents. First, results suggested that relative to rural areas, urban areas had a higher proportion of individuals living with paved roads (P < 0.001). Specifically, 59.09% of the rural residents lived with paved roads, and the figure surged to 74.00% among urban residents. In addition, a higher proportion of urban residents (41.66%) lived with a groundwater system, in comparison to their rural counterparts (20.18%, P < 0.001). Last, individual-level attributes, including household income, BMI, and physical activity, varied across urban and rural areas.

According to weighted analyses in Table 2, over 65% of study subjects had at least one NCDs, followed by 19.11% with two NCDs, 9.81% with three NCDs, and around 10% with at least 4 NCDs. The top three NCDs were arthritis or rheumatism (33.08%), hypertension (24.54%), and digestive disease (21.98%). In contrast, cancer, memory-related disease, and mental problems were the least prevalent NCDs. There were 0.90, 1.45, and 1.78% of the subjects self-reporting to have cancer, memory-related disease, and mental problems, respectively. Furthermore, urban subjects were more likely to have hypertension, heart problems, dyslipidemia, and diabetes, whereas rural subjects had a greater probability of asthma (P < 0.05). After controlling for CHARLS sampling weights, there existed no urban-rural disparities in the prevalence of each NCD since the 95% CI of urban residents overlapped with that of rural residents across all NCDs.

Table 3 presents results from negative binomial analyses. Results suggested that urban residents had a higher probability of more NCDs after controlling for all other covariates (Model 1 in Table 3, IRR = 1.05, 95% CI, 1.01–1.09). Results from Model 2, however, suggested that there existed no urban-rural disparities in the number of NCDs (IRR = 1.02, 95% CI, 0.95–1.02) after taking neighborhood-level variations into account. Results pertaining to urban vs. rural disparities in Model 2 were in line with those from subgroup analyses in Table 4.

Results in Table 3 suggest that living with paved roads was associated with fewer NCDs, in comparison with those living with unpaved roads. Specifically, after controlling for all covariates in Model 1, subjects living with paved roads (IRR = 0.87, 95% CI, 0.83–0.91) had a significantly lower incidence rate of more NCDs relative to those living with unpaved roads. Results here were consistent after controlling for neighborhood-level variations in Model 2. Subgroup analyses in Table 4 further suggested that in comparison with those living with unpaved roads, individuals living with paved roads were respectively less likely to have chronic lung disease (OR = 0.76, 95%CI, 0.63–0.92), chronic liver disease (OR = 0.74, 95% CI, 0.55–0.99), chronic kidney disease (OR = 0.67, 95%CI, 0.51–0.89), digestive disease (OR = 0.81, 95%CI, 0.69–0.96), arthritis or rheumatism (OR = 0.69, 95%CI, 0.55–0.87), and asthma (OR = 0.67, 95%CI, 0.50–0.88).

Living in neighborhoods with more primary care institutions was negatively associated with more NCDs (Model 2 in Table 3). Compared with those without primary care institutions in the neighborhood, individuals living with three primary care institutions or more were significantly less likely to have more NCDs (IRR = 0.94, 95% CI, 0.89–0.99). The disparities, however, did not exist after controlling for within-neighborhood variations.

Last, two individual-level attributes, household income, and BMI were associated with NCDs (Model 2 in Table 3). Specifically, compared with the poorest, the richest were less likely to suffer from a higher number of NCDs (IRR = 0.93, 95% CI, 0.88–0.98). Overweight (IRR = 1.14, 95% CI, 1.06–1.22) and obese subjects (IRR = 1.33, 95% CI, 1.21–1.46) respectively had a greater incidence rate of more NCDs compared with those with BMI less than 18.5. Subjects with BMI from18.5 to 25 had a lower incidence rate of more NCDs relative to those BMI less than 18.5, even after controlling for all other covariates (IRR = 0.90, 95%CI, 0.84–0.96).