Page 12 - JCAU-6-2

Journal of Chinese
Architecture and Urbanism Leisure-time physical activity

dependent variable. Cross-national comparative studies, other urban BE characteristics not explicitly categorized
research protocols, books or book chapters, dissertations, might impact LTPA levels, hence the inclusion of the term
reviews, editorials, and systematic reviews were also “other” to emphasize the potential diversity of influencing
excluded from the study. Subsequently, one of the authors characteristics. In addition, urban BE characteristics related
reviewed the abstracts of the selected articles, while another to “distribution” were not found in the included articles and
author with expertise in the subject area randomly reviewed were consequently excluded from our analysis.
abstracts to verify inclusion criteria. After these steps were
completed, the three authors reached a consensus on the 2.5. Coding evidence on the association between
final list of literature and jointly reviewed the full articles. urban built environment characteristics and
Table 1 provides a compilation of included articles. residents’ leisure-time physical activity levels
To analyze and compare the evidence of associations between
2.3. Extraction of study characteristics from the the seven urban BE characteristic categories mentioned
included articles in the previous subsection and residents’ LTPA levels, we
At least, two of the three authors examined the key coded the association evidence using statistical criteria from
details of the included articles, including the first author’s previous systematic reviews (Ding et al., 2011; Van Holle et al.,
name, year of publication, study design, city, sample 2012; Wendel‐Vos et al., 2007; Zapata-Diomedi & Veerman,
characteristics, methods for measuring the BE and 2016). Specifically, we coded the number of associations with
LTPA characteristics, BE characteristic extraction, LTPA the urban BE characteristic category that was significantly
measures, covariates, and other influencing factors. Table 1 associated with residents’ LTPA levels as a ratio of the number
displays the comprehensive details of the extracted study of positive, negative, and non-significant associations with
characteristics. In cases of disagreement between the two each characteristic category. Figure 1 and Table 3 present the
authors, the third author was consulted. This process flowchart illustrating the association encoding process and the
continued until all three authors reached a consensus. corresponding coding standards, respectively. First, an urban
BE characteristic category was considered to have convincing
2.4. Type classification of urban built environment evidence if more than 50% of the evidence showed the same
characteristics association with residents’ LTPA levels. Positive associations
Due to variations in the definitions and measurements were coded as “+,” while negative associations were coded
of urban BE characteristics among the included articles, as “-.” Second, when the evidence for the same association
the extracted data on urban BE characteristics differed between an urban BE characteristic category and residents’
significantly and were not amenable to direct quantitative LTPA levels ranged 40–50%, it was considered probable.
analysis. To address this limitation, we categorized different A probable positive association was coded as “(+),” and a
urban BE characteristics and explored their associations probable negative association was coded as “(-).” Third, if the
with residents’ LTPA levels using a framework of eight proportion of evidence for the same association was less than
BE categories proposed by existing studies (Giles-Corti 40%, or if a negative association was found in at least 25% of
et al., 2022; Giles-Corti et al., 2016). These categories the 40–50% evidence range, the association was considered to
include design, density, distance, destination accessibility, have “no evidence” and was coded as “0.” Conclusive testing
diversity, desirability, distribution, and demand required at least three articles. If this requirement was not
management. However, we replaced the expression met, the association was classified as “not applicable” and
“demand management” with “composite or other” in our coded as “N/A.”
classification. Table 2 details these classifications of the types
of BE characteristics. The term “demand management” 2.6. Determination of the quality of a study
specifically refers to traffic management measures and is not We developed our own criteria to assess the quality of the
directly related to urban BE characteristics associated with literature for this study, drawing commonly used literature
residents’ LTPA levels. Moreover, existing studies usually assessment tools in systematic reviews (Barnett et al.,
combine several urban BE characteristics into a single 2017; Cerin et al., 2017; Ewing & Cervero, 2010; Forsén
index for comparative analysis among cities. For example, et al., 2010; Grasser et al., 2013), in combination with the
the “community walkability index” assesses the extent Effective Public Health Practice Project (EPPHPP) tool for
to which an area is designed to support walking through assessing the quality of quantitative research (Gebel et al.,
urban design and is calculated based on the measurement 2015; Peters et al., 2020; Zapata-Diomedi & Veerman,
of several characteristics, such as density, diversity, design, 2016). Table 4 furnishes a comprehensive overview of the
destination accessibility, and distance to transit (Rundle et quality assessment criteria applied to the included articles.
al., 2019). Additionally, we considered the possibility that Our inclusion criteria were as follows: (i) achieving a

Volume 6 Issue 2 (2024) 3 https://doi.org/10.36922/jcau.2427