A Multi-Factor Framework for Cold-Climate Campus Design and Student Health

Abstract

This study explores how outdoor environments in cold-region university campuses influence students’ physical and mental health, addressing the lack of research on health-oriented campus design under cold climatic conditions. Drawing on Evidence-Based Design (EBD) theory and the Socio-Ecological Model (SEM), a Multi-Factor Analysis (MFA) framework integrating theoretical analysis, data mining, and empirical validation was developed to reveal the mechanisms linking campus environmental factors and student health. Through a systematic literature review and Latent Dirichlet Allocation (LDA) topic modeling, six key factors—climate adaptability, architectural layout, infrastructure, natural landscape, safety, and transportation accessibility—were identified and further verified through questionnaire data (N = 480) for reliability and validity. The Delphi method was then used to refine the indicator system and determine factor weights, while case studies of representative cold-region universities proposed optimization strategies from the dimensions of built environment, climatic adaptation, and perceived environment. The findings enrich the application of socio-ecological theory in health-oriented campus research and provide scientific and practical guidance for planning and promoting healthy university environments in cold regions.

1. Introduction

In 2016, the World Health Organization proposed the incorporation of land planning and landscape design into Healthy Cities construction policies, emphasizing the importance of formulating planning policies, health assessment, and monitoring [1]. The Ottawa Charter for Health Promotion explicitly proposed the goal of “Health for All” [2]. Youth represents a life stage in which university students are adapting to new environments and coping with multiple pressures, including academic stress and social challenges, making them more susceptible to physical and mental health issues [3]. Under cold climate conditions, these health problems may be further exacerbated, and an increasing number of common mental disorders, such as depression, anxiety, and stress, are becoming a global concern among young people and university students [4].

High-latitude regions are characterized by distinct seasonal variations, low annual average temperatures, and prolonged winter durations. Environmental characteristics of cold regions include extreme low temperatures, variable wind speeds, and reduced daylight hours in winter, which may significantly affect university students’ physical comfort, mood, and vitality [5]. Research has shown that the prevalence of overweight and obesity among university students continues to rise, which is closely associated with unhealthy dietary patterns and lack of physical activity [6]. Particularly in urban universities, convenience foods and sedentary study behaviors have become common phenomena. Jayman et al. found a significant association between lack of physical activity and chronic fatigue syndrome and metabolic diseases [7]. The primary factor affecting health is the lack of physical exercise, and in cold regions, there are multiple contributing factors. Extreme low temperatures, strong winds, and heavy precipitation create difficulties for students and faculty to exercise on campus, such as in cycling activities [8]. These challenges are compounded by negative impacts from university-specific environmental factors. University campuses built on a large scale in China during the early 1990s, despite featuring grand architectural styles and designs, lack refined environmental construction from health and climate perspectives [9]. Furthermore, winter conditions lead university students to become overly dependent on motorized transportation, virtual interactions, and food delivery services, cultivating sedentary lifestyles. This has resulted in a significant decline in physical activity levels among Chinese adults over the past 20 years, with this decline being particularly pronounced in cold climates and high-latitude regions [10,11].

Individual health and well-being are comprehensively influenced by multi-level environmental factors, including climatic conditions, spatial organization, social interactions, and perceptual experiences [12]. Although academia has increasingly emphasized the relationship between environment and health in recent years, most studies remain focused on single dimensions, such as building thermal comfort, green space ratios, or behavioral and psychological responses [13,14], lacking cross-level and cross-system integrated perspectives. Particularly in cold climate contexts, the combined effects of extreme climatic conditions and enclosed spatial configurations make university campuses an ideal setting for examining the interactive mechanisms between environment and health.

Meanwhile, existing health-oriented design frameworks, such as the WELL Building Standard, healing environment models, and biophilic design theory, provide important references for health-promoting spatial design. However, these models primarily focus on optimization at the micro-spatial and perceptual levels, with insufficient attention to how multi-level socio-ecological mechanisms jointly influence health outcomes [15]. Furthermore, existing Post-Occupancy Evaluation (POE) and environmental psychology research predominantly concentrate on indicators such as satisfaction, thermal comfort, or environmental preferences, lacking systematic studies that integrate climate adaptability, spatial behavior, and perceptual experiences within a unified analytical framework [16,17]. Based on this, the present study combines the theoretical foundations of Evidence-Based Design (EBD) and the Socio-Ecological Model (SEM) to propose a Multi-Factor Analysis Framework (MFA), which systematically reveals the interactive mechanisms among the three-dimensional elements of built environment, climatic environment, and perceptual environment on students’ physical and mental health in university campuses in cold regions, thereby providing theoretical support and practical references for health-oriented campus planning and design.

1.1. Theoretical Foundation

1.1.1. Evidence-Based Design Theory

Evidence-Based Design (EBD) theory originates from the principles of evidence-based medicine. In the context of the built environment, it refers to “a process in which designers collaborate with clients to carefully examine and analyze the most reliable scientific research evidence available to make informed design decisions” [18,19]. EBD represents both a design paradigm emphasizing the procedural nature of design and a methodological approach stressing the scientific foundation of design decisions. The introduction of this concept addresses the overly subjective tendencies in traditional built environment design practices, ensuring stronger alignment between design outcomes and project objectives in design processes with clearly defined goals [20].

1.1.2. Socio-Ecological Model

Since the 1950s, the mechanisms underlying health behavior formation have become a central research focus across multiple disciplines, generating various theories and models aimed at identifying key determinants of individual behavior [21]. However, traditional behavioral models have predominantly focused on single dimensions, such as individual cognitive attitudes or social support systems [22,23], failing to capture the dynamic interactions between individuals and their surrounding environments. To address these limitations, Sallis et al. developed the Ecological Model of Behavior based on socio-ecological theory, emphasizing that health behaviors are shaped by interactive influences of multi-level and multi-system environmental factors [24]. This model categorizes environmental influences into several interrelated layers—ranging from individual and interpersonal factors to broader social structures, natural and built environments, perceptual and informational environments, behavioral settings, and public policies—forming a multidimensional framework of behavioral determinants [25,26]. Importantly, these layers are not linearly connected through cause-effect chains but operate through dynamic and reciprocal interactions, wherein behavioral mechanisms serve as critical mediators linking environment to health outcomes [12].

Compared to traditional models, the socio-ecological model is distinguished by its comprehensiveness and systematicity. It considers not only micro-level individual psychological and interpersonal factors but also incorporates macro-level structural determinants such as institutional policies and physical spaces. By revealing the interactive mechanisms among different levels, it provides a more comprehensive explanation of the pathways through which behaviors are generated and changed [27]. Under the guidance of this model, researchers can transform abstract environmental dimensions into operationalizable research indicators. For instance, “built environment” can be represented by physical spatial elements such as campus green space coverage, walkability, and outdoor recreational facility density [28]; “climatic environment” can be measured through meteorological variables such as temperature, snowfall, and wind chill index [29], while “perceptual environment” can be assessed through methods such as questionnaire surveys to evaluate students’ subjective experiences of safety, comfort, spatial attractiveness, and usability [30].

Within the specific context of university campuses in cold regions, extreme climatic conditions and spatial enclosure significantly influence students’ daily activity patterns [31]. Based on the Ecological Model of Behavior, the present study extracts three key dimensions closely related to university students’ outdoor behavior—built environment, climatic environment, and perceptual environment—aiming to systematically explore how campus environments influence students’ physical and mental health through behavioral mechanisms, thereby providing theoretical support and practical references for environmental improvement and health promotion in cold-region higher education institutions [32].

1.2. Research Objectives

Is it possible to mitigate the adverse effects of cold climates on students by optimizing campus environments? This study systematically identifies the key factors affecting student health in university campuses located in cold regions. This objective is achieved through three sub-goals:

(1)

To comprehensively identify and analyze the key determinants influencing student health within university campuses in cold regions of China.

(2)

To distinguish and evaluate the most significant environmental factors that have measurable impacts on student health in these contexts.

(3)

To propose specific guidelines for optimizing the physical environment of university campuses in cold climates, aiming to mitigate the negative health impacts of cold weather, enhance the health-supportive functions of campus environments, and provide scientific evidence and practical guidance for future planning and policy-making in related fields.

2. Methods

2.1. Research Framework

This study adopts the principles of Evidence-Based Design (EBD) and the Socio-Ecological Model (SEM) to develop a Multi-Factor Analysis Framework (MFA) for systematically evaluating the pathways through which the physical environment of university campuses in cold regions impacts student health. The framework integrates three sources of information: literature-based data mining, expert consensus evaluation, and validation through representative case studies, emphasizing a unified approach combining theoretical grounding, methodological rigor, and practical applicability [18]. To ensure the scientific validity and practical relevance of the findings, the framework adopts the following three-phase strategy.

2.1.1. Extraction of Key Factors Through LDA Topic Modeling

Given the multidimensional nature of health—encompassing physical, psychological, and social dimensions—it is difficult to directly and systematically quantify health outcomes in textual sources such as literature and case documents. Therefore, this study adopts “student outdoor activity” as a proxy indicator for health. Within the socio-ecological framework, student activity is conceptualized as a mediating mechanism linking campus environment to health outcomes: the environment first influences student activity, which in turn affects health. Accordingly, Latent Dirichlet Allocation (LDA) topic modeling was employed to extract frequently co-occurring environmental theme terms (e.g., “green space,” “path accessibility,” “spatial safety,” “social facilities”) from multi-source textual data, thereby identifying key environmental factors that influence student activity behavior. Given the well-established positive correlation between student activity and health demonstrated in existing research, these environmental themes influencing activity can also be interpreted as critical determinants of student health. Compared to traditional variable operationalization approaches, this data-driven, bottom-up method allows for the discovery of influential factors without predefined indicators, providing both theoretical and methodological support for the integrated analysis of heterogeneous data sources.

2.1.2. Expert Evaluation via Delphi Method

Building on the findings from Phase I, 20 senior experts from the fields of architectural design, urban planning, environmental psychology, and public health were invited to evaluate the scientific relevance and importance of the preliminary environmental indicators using the Delphi method, establishing a prioritized ranking system for the indicators [19].

2.1.3. Validation Through Representative Campus Case Studies

In the final phase, representative university campuses were selected as research cases for on-site investigation to examine how the identified environmental factors are manifested in actual campus spaces. Through contextualized validation, this phase ensured that the proposed design guidelines possess practical feasibility and optimization value in real spatial contexts [33].

Ultimately, the study developed a health-oriented campus design optimization system based on a three-dimensional environmental classification, comprising five primary dimensions and nine secondary indicators, along with a series of actionable strategies. This framework provides both theoretical support and methodological pathways for constructing health-supportive campus environments in cold-region universities (Figure 1).

3. Results and Discussion

3.1. LDA Topic Modeling

This study employed Latent Dirichlet Allocation (LDA) topic modeling to extract and analyze key themes from literature related to the impact of cold-region university campus environments on student health. The process consisted of four stages: literature screening, data preprocessing, LDA topic modeling, and data analysis. First, the PRISMA protocol was followed to conduct comprehensive literature retrieval, screening, and quality assessment. Second, the Jieba tokenizer, version 0.42.1, was used for word segmentation, and stop words were removed to optimize the textual dataset. The Term Frequency-Inverse Document Frequency (TF-IDF) method was applied to enhance term distinctiveness. Subsequently, the Gensim library, version 4.3.2, was employed to construct the LDA topic model, generating both “topic-word” and “document-topic” probability distributions. The model’s coherence score was used to determine the optimal number of topics. Finally, pyLDAvis, version 3.4.1, was employed for interactive topic visualization to ensure clear thematic distinctions.

The core environmental factors extracted via LDA modeling formed the foundation of the study’s Multi-Factor Analysis Framework (MFA), revealing that university campus environments in cold regions impact student health across multiple dimensions, including climate adaptability, architectural typologies, facility design, natural and green spaces, safety, and campus mobility. These findings ensure that subsequent research can comprehensively explore how multi-level environmental optimization strategies influence student health outcomes.

3.1.1. Literature Screening

This study established the research scope of “cold-region university campuses worldwide” through the retrieval, screening, and analysis of relevant literature. The research adopted the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework for systematic literature retrieval and screening [34]. The literature search was conducted from July to October 2024, with data sourced from Google Scholar, Scopus, ScienceDirect, ProQuest, ResearchGate, and CNKI, covering the period from 2000 to 2024. Only peer-reviewed journal articles published in English and Chinese were included. Keywords such as “cold region,” “cold climate,” “cold city,” “physical factors,” “environmental factors,” “university campus,” “outdoor environment,” and “health” were used, with Boolean logic (‘OR’ and ‘AND’) applied for combination optimization, resulting in an initial pool of 1943 relevant studies.

During the initial screening phase, inclusion and exclusion criteria were established based on PRISMA standards [35]. The study retained literature focusing on students’ outdoor activity behaviors during cold seasons and their associated health impacts, such as physical activity, perceptual experiences, and thermal comfort. Only peer-reviewed journal articles published after the year 2000 were included. Duplicate and irrelevant studies were removed, resulting in a selection of 50 articles. To supplement the screening process, the research team employed reference snowballing and expert recommendations [36], leading to the inclusion of an additional 10 relevant studies. Ultimately, 60 papers were finalized for analysis (Figure 2).

For quality assessment, the Crowe Critical Appraisal Tool (CCAT) [37] was employed to evaluate the methodological rigor of the selected studies, with reference to subsequent validations of the tool’s reliability. Two research members independently assessed the CCAT scores and reached consensus through discussion. A five-point Likert scale was used for scoring, and only articles with a quality rating of 65% or higher were retained in the final dataset [38].

3.1.2. Textual LDA Topic Discovery

Data Cleaning

In LDA topic discovery, data preprocessing is a crucial step to ensure both model accuracy and interpretability, with word segmentation and stop-word removal being the most essential components [39]. Literature-derived textual data contain a large number of numerals, punctuation marks, and semantically meaningless Chinese words. These noisy elements are not only irrelevant to the thematic content but also interfere with machine interpretation [40]. Therefore, this study first imported a domain-specific lexicon relevant to the research topic and employed the Jieba library in Python version 3.10, for word segmentation [41]. Segmented text better captures semantic structures, thereby improving the precision and interpretability of the topic model. Furthermore, to remove frequently occurring but thematically irrelevant words, a stop-word list containing 1482 entries was applied to eliminate general function words and conjunctions. This step significantly enhanced both the computational efficiency of the model and the interpretability of the resulting topics [42].

LDA Topic Modeling and Parameter Optimization

Latent Dirichlet Allocation (LDA) is an unsupervised text mining method used to identify latent topics within large-scale corpora [43]. pyLDAvis is a visualization tool that intuitively displays the relationships among different topics [44]. In this study, the optimal number of topics for analyzing environmental factors influencing student health in cold-region university campuses was determined using two complementary approaches: perplexity analysis and visual inspection.

The number of topics is a key parameter in topic modeling. This study employed the perplexity metric to determine the optimal number of topics (k). Perplexity reflects the degree of uncertainty in topic assignment across documents. When the decreasing trend of perplexity begins to level off or an inflection point appears, the corresponding k value is considered optimal [45]. As shown in Figure 3, the average perplexity decreases as the number of topics increases. Based on this trend, the optimal number of topics was determined to be six (k = 6).

In addition, the results of the LDA model were visualized using pyLDAvis. When the number of topics was set to six (num_topics = 6), each circle in the visualization represented a distinct topic. The size of each circle indicated the proportion of documents associated with that topic, while the distance between circles reflected the semantic similarity between topics [46]. As shown in Figure 4, the six topics appear clearly separated with no significant overlap, supporting the conclusion that six is the optimal number of topics.

3.1.3. LDA Topic Data Analysis

Data Results

The preprocessed abstract texts were used to train the LDA model via the Gensim library in Python. Based on the previously determined results, the number of topics (K) was set to 6. The model parameters were configured as α = 1/K = 0.16, β = 0.01, with 20 iterations. Through training on word frequency and latent semantic distributions within the corpus, the model extracted the seven words with the highest probability for each topic [48,49]. These high-probability keywords served as descriptors to characterize topic semantics, and combined with semantic features, enabled topic identification and labeling (

Table 1. Multilevel Thematic Analysis Results of Campus Environment and Student Health in Cold Regions (Based on LDA Modeling).

Topic	Theme	Core Keywords	Pathway Mechanism (Mechanism Level)	Mechanistic Interpretation (Social-Ecological Level)	Socio-Ecological Perspective
Topic 0	Climatic Conditions	Low temperature, snowfall, strong wind, sharp temperature variation, wind speed change, ice-snow utilization, dryness	Winter resource integration and utilization; outdoor microclimate regulation; protection design for extreme cold	Climatic Environment Level (Environmental Stress Theory, Climate–Health Model)	Climatic Environment
Topic 1	Architectural Form	Building layout, daylighting design, orientation, insulation, building materials, corridor connection	Spatial organization and scale control; thermal-light balance	Built Environment Level (Building–Environment–Health Model)	Built Environment
Topic 2	Campus Infrastructure	Anti-slip pavement, resting facilities, temporary shelters, walking paths, covered corridors, ramps, trash bins	Pedestrian facility quality; path obstacle avoidance; path interface design	Built Environment Level (Health-Supportive Environment Theory)	Built Environment
Topic 3	Natural Environment and Green Space	Outdoor green classrooms, cold-resistant plants, seasonal vegetation, ice sculptures, biodiversity support, windbreak plants	Seasonal vegetation configuration; ecological protective design	Perceptual Environment Level (Biophilia and Restorative Environment Theories)	Perceived Environment
Topic 4	Safety Factors	Icy roads, pathway maintenance, safety perception, lighting, surveillance, gender difference, anti-slip design	Traffic safety; crime prevention and safety; winter activity protection	Perceptual Environment Level (Safety Perception and Risk Cognition Models)	Perceived Environment
Topic 5	Transportation and Accessibility	Transportation connection, public transport, pedestrian paths, bicycle lanes, parking management, walking network, tactile paving	Pedestrian system continuity; transport diversity	Built Environment Level (Active Transport and Accessibility Equity Theories)	Built Environment

Note: Each topic corresponds to a distinct level within the social–ecological framework: the Climatic Environment Level (macro level), the Built Environment Level (meso level), and the Perceptual Environment Level (micro level). Together, these levels constitute the integrated logic of climate exposure–spatial intervention–health response.

).

Additionally, the “document–topic” distribution generated by the LDA model is shown in

Table 2. Document-Topic Probability Distribution in the LDA Model.

Document ID	Topic 0	Topic 1	Topic 2	Topic 3	Topic 4	Topic 5	Associated Topic
0	0.750125	0.085132	0.0654185	0.0513375	0.0458125	0.0021745	0
1	0.402548	0.075189	0.092354	0.352547	0.014865	0.062497	0
2	0.065485	0.525486	0.254867	0.075156	0.023545	0.055461	1
3	0.048456	0.085481	0.804656	0.016245	0.025485	0.019677	2
4	0.025489	0.085425	0.258949	0.0451896	0.012545	0.5724024	5
......	......	......	......	......	......	......
55	0.155685	0.002545	0.054568	0.154588	0.025458	0.607156	5
56	0.065844	0.045264	0.098545	0.526545	0.125996	0.137806	3
57	0.025645	0.095465	0.154898	0.012655	0.648584	0.062753	4
58	0.751522	0.015456	0.054521	0.015458	0.054521	0.108522	0
59	0.025486	0.125485	0.545664	0.094521	0.05553	0.153314	3

, presenting the distribution probability of different topics within each text. Higher probability values indicate stronger thematic relevance, allowing for rapid identification of the most prominent topic content within each text [50].

Textual Topic Analysis of Campus Environment and Student Health in Cold Regions

Based on the topic identification results of the LDA model, this study further conducted a systematic interpretation of the six themes from a socio-ecological perspective. Guided by socio-ecological theory, the analysis adopts a multi-level approach encompassing the macro-environment, meso-spatial, and micro-individual dimensions to systematically organize the semantic cores, functional pathways, and mechanism characteristics of each theme. Through an analytical logic of “semantic core–mechanistic pathway–theoretical mapping,” a comprehensive framework was constructed connecting natural exposure, spatial intervention, and health response, revealing the internal logic and hierarchical progression of environmental health effects in cold-region universities. The relevant results are presented in Table 1.

• Climatic Conditions and Student Health (Topic 0)

The climatic characteristics of cold regions are primarily manifested in low temperatures, heavy snowfall, strong winds, and sharp temperature fluctuations, all of which directly affect students’ physiological comfort and psychological well-being. Studies have shown that the thermally neutral PET value for university students in high-altitude cold regions of China during winter is approximately 12.1 °C [51], indicating that climatic exposure in cold regions has a significant impact on student behavior and comfort perception. Moreover, wind speed, ground temperature, and radiative variation are considered key variables influencing thermal sensation [52]. From the macro-environmental perspective of socio-ecological theory, such climatic exposure constitutes an external stressor to student health, whose effects are progressively mitigated and transformed through three primary pathways: winter resource integration, microclimate regulation, and protective design.

Within this adaptive framework, winter resource integration embeds ice and snow resources into campus culture and outdoor activity systems—such as ice sculpture art and snow sports—transforming a passive defense strategy into active climate utilization, thereby turning cold environments into positive mediators for social engagement and psychological restoration. Second, through optimization of campus form and landscape organization, including shading design, wind-direction guidance, and thermal radiation balance, outdoor thermal comfort is enhanced to sustain the continuity of winter activities. Protective design measures, such as windbreak belts and heated pavements, effectively reduce wind-chill indices and slipping risks, providing safer and more stable travel conditions for students.

This transformation—from resistance to adaptation and ultimately utilization—illustrates the dynamic mechanism of the climate–health relationship. It aligns with environmental stress theory [53], which emphasizes the “external stimulus–adaptive response” regulatory logic, and with the climate–health framework [54], which reveals that cold-region campuses are not merely passive recipients of climatic exposure but rather co-shape health responses through spatial design and behavioral strategies.

• Architectural Form and Student Health (Topic 1)

Building on the foundation of addressing climatic exposure, architectural form serves as a core mediator for environmental regulation in cold-region campuses, enhancing the buffering capacity against cold-weather impacts. The layout, daylighting, orientation, and insulation performance of campus buildings collectively determine students’ thermal environment, luminous environment, and psychological comfort. Typically, campus buildings in cold regions adopt enclosed or corridor-connected configurations to minimize cold air convection, form “warm cluster effects,” and reduce outdoor exposure time for faculty and students while stabilizing the local thermal environment.

This health-oriented architectural strategy operates through three interrelated pathways: spatial layout and organization, scale and thermo-luminous balance, and insulation and material performance. Compact building cluster layouts and organization can reduce cold exposure and optimize microclimate conditions, enabling buildings to create localized thermal self-circulation at the spatial level. Appropriate window-to-wall ratios and building depth establish a balance between energy consumption and natural daylighting, thereby maintaining comfort simultaneously in both visual and physiological dimensions. Meanwhile, high-insulation wall systems and hollow glass windows reduce heat loss while maintaining air circulation, helping to prevent sick building syndrome.

Together, these three mechanisms not only optimize the physical environment but also shape environmental perceptibility and psychological safety at the psychological level—echoing the view in environmental psychology that spatial perception influences restorative experiences [12]. Simultaneously, architecture assumes an active role in health regulation throughout this process. Its morphological and performance design embodies the “Design as Health” principle [55], transforming cold-region architecture from a purely defensive shell into a built ecological unit that promotes human well-being.

• Campus Infrastructure and Student Health (Topic 2)

In cold-region university environments, infrastructure serves as a critical layer linking spatial function with health behavior, playing a pivotal role in moderating environmental stress and supporting daily activities. Features such as anti-slip pavements, sheltered spaces, covered walkways, and pedestrian systems not only determine the safety and convenience of mobility but also influence students’ willingness to be active and their psychological states in harsh climatic conditions. From the meso-spatial perspective of socio-ecological theory, infrastructure functions as a mediating element between the built environment and behavioral health, with its mechanisms primarily reflected in three aspects: pedestrian facility quality, pathway continuity, and interface experience.

High-quality pedestrian infrastructure forms the foundation for safety and health. Material technologies such as anti-slip surfacing, heated pavements, and permeable concrete effectively reduce the risk of falls and maintain surface friction stability during snow conditions. This process not only mitigates external physical stress but also enhances a sense of psychological security. Second, pathway continuity design—achieved through the integration of barrier-free ramps and covered corridors—prevents interruptions and exposure risks, ensuring that campus mobility remains accessible and efficient even during cold winter seasons. The concept of “behavioral path predictability” [56] and the coherent spatial logic align with environmental behavior theory, helping to maintain students’ travel confidence and physical activity levels. Moreover, detailed interface design along pathways—including lighting, warning signage, and greenery configuration—provides visual guidance while creating an ambient environment that combines warmth and order, thereby promoting psychological comfort and social interaction at the perceptual level.

Together, these three aspects transform campus infrastructure from a mere transportation system into a health-supportive environment. This logic resonates with the Health Promotion Model [57] and Walkability Theory [58], illustrating how the built environment, through spatial organization and behavioral guidance, jointly shapes health outcomes through meso-level regulatory mechanisms.

• Natural Environment and Green Space (Topic 3)

Beyond physical environmental regulation, natural elements play an indispensable role in providing psychological support within cold-region campuses. As an emotional extension of the infrastructural system, the natural environment not only fulfills ecological regulatory functions but also possesses psychological healing and social interaction value. From the micro-level perspective of socio-ecological theory, natural features alleviate environmental stress through both visual and emotional pathways, primarily via seasonal vegetation design, ice-and-snow landscape utilization, and ecological protection design.

Appropriate seasonal vegetation composition—combining evergreen and deciduous species—maintains visual greenery in cold-region campuses during winter, reducing desolation and mood fluctuations. Visual continuity aligns with the attention restoration and emotional recovery mechanisms emphasized in Restorative Environment Theory [13]. Furthermore, creative utilization of ice and snow landscapes—such as ice sculpture art zones, snow activity fields, and temporary ice-and-snow exhibitions—transforms ice and snow from climatic burdens into aesthetic and social assets, promoting positive experiences and psychological resilience among students. Meanwhile, the integrative logic of cold-region ecological design theory [59] supports ecological protection design through windbreak belts, topographical greenery, and ecological slopes, which not only improve local microclimate conditions but also provide quiet and sustainable outdoor study and leisure spaces. The interaction among these three mechanisms enables the natural environment to form a dual support system for both physiological and psychological well-being.

• Campus Safety Factors and Student Health (Topic 4)

In cold climates, campus safety encompasses both physical and psychological dimensions. A lack of safety perception stems not only from physical hazards caused by ice and snow but also from psychological anxiety arising from insufficient nighttime lighting, surveillance blind spots, and environmental unfamiliarity. The safety system therefore constitutes an essential component of the campus health framework. From the micro-level perspective of socio-ecological theory, the safety mechanism in cold-region universities operates through three interrelated pathways: traffic safety protection, nighttime safety perception enhancement, and winter activity protection.

Considering the theoretical pathway of Environmental Safety Perception Theory [60] that “physical order shapes psychological stability,” traffic safety protection forms the foundational level. The application of anti-slip paving, ground heating, and snow removal systems effectively reduces walking risks and makes winter mobility more controllable. The cultivation of nighttime safety perception depends on the rational layout of lighting and surveillance systems, which eliminate visual blind spots and improve lighting continuity, thereby reducing gender-related safety anxiety and making campus spaces more psychologically inclusive. Meanwhile, well-developed winter activity protection systems—including emergency equipment, monitoring systems, and emergency response training—not only strengthen students’ sense of environmental control but also enhance trust and collective security at the social level. Together, these three pathways construct a progressive structure of “physical safety–psychological safety–health promotion,” aligning with the Risk Perception Model [61] and reflecting the safety system as a vital node within the perceived-environment dimension of the socio-ecological network.

• Transportation and Accessibility (Topic 5)

The transportation system forms the backbone of campus spatial networks, with accessibility and modal diversity directly determining students’ physical activity levels and everyday well-being. In cold-region universities, harsh weather often limits mobility and reduces activity levels; therefore, transportation planning serves not only functional efficiency but also serves as a crucial pathway for health promotion. From the meso-spatial perspective of socio-ecological theory, campus transportation affects health through three primary mechanisms: pedestrian system continuity, transportation diversity, and travel experience optimization.

A continuous pedestrian system—through proper control of pathway width, slope, and slip resistance—ensures safety while increasing students’ activity frequency, maintaining basic physical metabolism and social contact. This perspective validates the Active Transportation Theory [62], which emphasizes the principle that “travel is activity.” In terms of transportation diversity, the coordinated layout of bus routes, bicycle lanes, and pedestrian networks reduces dependency on motor vehicles, promoting low-carbon and inclusive mobility patterns. This improvement in equity reflects the emphasis on “equality in travel rights” highlighted by the Accessibility Justice Model. Furthermore, through the design of covered walkways, semi-open rest nodes, and heated corridor systems, travel experience gains comfort extension under climatic constraints, enabling mobility behavior to transcend seasonal limitations and forming a positive feedback loop between environmental adaptation and social vitality.

Collectively, these mechanisms demonstrate that the transportation system is not merely an isolated circulation structure but a vital health infrastructure within the built-environment layer of the socio-ecological framework, connecting the spatial, behavioral, and psychological dimensions.

This study analyzes the multidimensional mechanisms through which campus environments in cold regions affect student health across six dimensions. The findings reveal that climatic conditions, architectural forms, infrastructure, natural environments, safety systems, and transportation accessibility jointly constitute the key environmental factors influencing student health. These factors act on individual health outcomes through the triple pathways of space–behavior–psychology, forming a dynamic and systematic chain of influence.

At the theoretical level, this chapter deepens the logical chain of “environmental stress–behavioral regulation–health outcomes,” providing systematic theoretical support for health-oriented campus development in cold-region universities. At the practical level, it suggests that universities should strengthen climate adaptation, ecological integration, and health-oriented design strategies in campus planning and management, achieving a transformation from passive defense to proactive health promotion. Based on the above analysis, an overview of the LDA thematic analysis results on the environmental–health relationships in cold-region campuses is presented in Table 1.

3.1.4. Empirical Validation Results of the Structural Framework of LDA-Derived Influencing Factors

To verify the structural validity of the framework, a questionnaire entitled “Impacts of Cold-Region University Campus Factors on Student Health (Validation Version)” was developed based on the LDA-derived model. The questionnaire was distributed via the Wenjuanxing platform to students from four universities in Heilongjiang Province, and a total of 480 valid responses were collected. The results of the reliability and validity analysis for the framework are shown in

Table 3. Validation Results of the Reliability and Structural Validity of the Framework on the Impacts of Cold-Region University Campus Factors on Student Health.

Validation Indicators	Primary Dimension	Secondary Dimension
Cronbach’s α Coefficient	0.946	0.977
Kaiser–Meyer–Olkin (KMO) Measure	—	0.952
Bartlett’s Test of Sphericity	—	p < 0.001
Cumulative Variance Contribution Rate	—	92.290% (Six factors extracted)
Number of Items	6	13
Valid Cases (N)	480

.

Cronbach’s α coefficient, which is commonly used to assess the internal consistency of questionnaires, was employed to test the structural reliability of the framework [63]. The Cronbach’s α values for both hierarchical levels exceeded 0.90, indicating that both dimensions of the framework possess good reliability and internal consistency [64]. The Kaiser–Meyer–Olkin (KMO) statistic was used to evaluate whether the data were suitable for factor analysis, thereby indirectly reflecting structural validity [65]. The results showed that the KMO value for the 15 secondary dimensions was 0.952 (>0.900), indicating excellent sampling adequacy and suitability for factor analysis. Bartlett’s test of sphericity for the second-level dimensions yielded p < 0.001, suggesting significant correlations among variables and confirming that the data structure was appropriate for factor analysis. Exploratory Factor Analysis (EFA) results showed that six factors were extracted from the 15 secondary dimensions, with a cumulative variance contribution rate of 93.921%. This indicates that the extracted factors effectively capture the underlying structure and relationships of the original data, thereby confirming that both hierarchical dimensions of the framework possess good structural validity [66].

3.2. Collection and Analysis of Expert Opinions Using the Delphi Method

3.2.1. Expert Selection for the Delphi Method

The purpose of the Delphi method is to obtain the most reliable consensus among a panel of experts through a structured process of iterative questionnaire feedback [67,68]. In this study, the Delphi method was employed to validate and refine the key physical factors identified through LDA topic modeling, thereby ensuring the scientific rigor of the indicator system [69].

A non-probabilistic purposive sampling approach was adopted. In October 2024, twenty consulting experts were invited from universities, research institutions, and architectural and landscape design firms. Each expert was carefully selected based on the following criteria: research, technical, or managerial personnel with at least 10 years of experience in cold-region urban environmental design (including architecture, urban design, landscape design, and related fields), university campus design, or health behavior promotion. In addition, experts were required to possess proficient domain knowledge, professional competence, and familiarity with interdisciplinary perspectives [70]. The general profile of the expert sample is presented in

Table 4. Basic Information of Expert Samples.

Years of Professional Experience			Academic Title			Educational Qualification		Nationality
10–20 Years	20–30 Years	Over 30 Years	Lecturer	Associate Professor	Professor	Master’s Degree	Doctoral Degree	Non-Chinese	Chinese
5	8	7	2	11	7	8	12	0	20

.

In the first round of consultation, experts rated six primary indicators and fifteen secondary indicators using a five-point Likert scale and conducted self-assessments of their own authority levels. The research team then calculated statistical parameters such as mean values, standard deviations, and coefficients of variation (CV) to preliminarily adjust the indicator system, consistent with common statistical analysis methods in Delphi studies [71]. In the second round, based on feedback from the first round, experts were asked to assign weights to each indicator (totaling 100 points). Ultimately, six primary indicators and nine secondary indicators were finalized. A criterion of CV < 0.25 was adopted to determine consensus among expert opinions, providing a theoretical foundation for subsequent case observations.

3.2.2. Development Tools

In the first round of consultation, the “Consultation Questionnaire on the Impact of Cold-Region University Environments on College Students’ Health (Part I)” was distributed to experts. The questionnaire required experts to evaluate the preliminary indicators of environmental factors influencing students’ health. Experts were asked to rate each factor in terms of its importance, basis of judgment, and degree of familiarity using a quantitative coding system. This scoring method, following established research [72], converted qualitative judgments (e.g., “very important,” “familiar”) into numerical scores (e.g., 10, 8, 6, 4, 0), facilitating the calculation of the expert authority coefficient and consistency analysis (see

Table 5. Quantitative composition of the indicators.

Importance	Quantitative Value	Basis for Judgment	Quantitative Value	Proficiency Level	Quantitative Value
Very Important	10	Practical Experience	0.8	Very Familiar	10
Important	8	Theoretical Analysis	0.6	Familiar	8
Somewhat Important	6	Peer Understanding	0.4	Somewhat Familiar	6
Moderately Important	4	Intuition	0.2	Slightly Familiar	4
Not Important	0			Not Familiar	0

).

The questionnaire consisted of three sections: The first section introduced the research background and briefly explained the Delphi method’s operational process and its role within the overall study. In this section, experts were asked whether they agreed to participate in two rounds of consultation, and their basic demographic and professional information was collected. Reporting such information is essential for assessing the credibility of the expert panel [73]. The second section, based on this study’s interpretation of six environmental indicators, asked experts to rate their level of agreement using the five-point Likert scale. The third section required experts to conduct a self-assessment of their authority level in this round of consultation. After the first consultation, expert feedback was compiled, and statistical metrics—including the mean, standard deviation, full-score rate, coefficient of variation (CV), expert enthusiasm coefficient, and authority coefficient—were calculated to revise and optimize the indicator system based on the statistical results.

In the second round of consultation, the revised version of the “Consultation Questionnaire on the Impact of Cold-Region University Environments on College Students’ Health (Part II)” was distributed to the same expert panel. Experts were presented with the modified indicators and asked to assign weights to each environmental factor according to its dimension, with a total score of 100 points. The weights for each dimension were ultimately determined based on expert opinions. In addition, experts were asked again to self-assess their authority level. After this round, statistical indicators—including the mean, standard deviation, coefficient of variation, expert enthusiasm coefficient, and authority coefficient—were computed for the weights assigned to each dimension to establish the indicator system of environmental factors influencing college students’ health.

The questionnaires were distributed via email. Based on the designated panel members, a total of 20 expert questionnaires were sent, and 16 valid responses were received, yielding an effective response rate of 80%. In the second round of the Delphi procedure, 5 experts revised their opinions based on the results from the first round, while 11 experts maintained their previous judgments.

In the Delphi method, the mean, standard deviation, full-score rate, and coefficient of variation (CV) are the core statistical indicators for analyzing expert opinions. A higher mean generally reflects a higher perceived importance of a given issue by the experts, which is particularly crucial in decision-making processes [74]. A smaller standard deviation indicates that experts have reached consensus on an issue, with a standard deviation below 0.5 typically indicating that consensus has been achieved. The full-score rate serves as a supplementary indicator of expert opinion consistency, and when it reaches or exceeds 70%, it usually signifies high consistency among experts [75]. A low coefficient of variation—typically less than 0.25—indicates that experts’ opinions on an issue are relatively consistent, suggesting that preliminary consensus has been reached [72]. For example, indicators with a coefficient of variation below 0.2 are generally included in the priority list.

3.2.3. Data Analysis and Results

Statistical analysis was conducted using SPSS 22.0 software. In the first round of expert consultation, indicators with a mean score ≥ 4 and a full-score rate ≥ 80% were retained, resulting in six primary indicators and fifteen secondary indicators. The mean values of all indicators ranged from 4.12 to 4.60, the standard deviations ranged from 0.49 to 0.88, and the coefficients of variation were between 0.11 and 0.21, indicating a high level of expert agreement with the preliminary indicators and strong consistency in opinions (

Table 6. Results of the First Round of Expert Consultation Analysis.

Environmental Factors	Mean	Standard Deviation	Full Score Rate	Coefficient of Variation	Secondary Indicator	Mean	Standard Deviation	Full Score Rate	Coefficient of Variation
Climate	4.36	0.84	82.40%	0.20	1. Winter Resource Integration and Utilization	4.40	0.50	88.00%	0.11
					2. Microclimate Regulation for Outdoor Activities	4.36	0.49	87.20%	0.11
					3. Protective Design for Extreme Climatic Conditions	4.40	0.50	88.00%	0.11
Architectural Forms	4.40	0.81	81.20%	0.19	4. Spatial Layout and Organization	4.60	0.50	92.00%	0.11
					5. Spatial Scale	4.56	0.50	91.20%	0.11
Campus Facility Characteristics	4.32	0.84	87.30%	0.16	6. Pedestrian Facility Quality	4.36	0.64	87.20%	0.13
					7. Pathway Obstacle Avoidance	4.44	0.51	88.80%	0.11
					8. Pedestrian Pathway Interface Design	4.44	0.65	88.80%	0.13
Natural Environment and Green Spaces	4.12	0.82	87.20%	0.17	9. Seasonal Vegetation Configuration	4.36	0.57	87.20%	0.13
					10. Utilization of Snow and Ice Resources	4.28	0.58	85.60%	0.14
Campus Safety	4.46	0.84	88.00%	0.19	11. Traffic Safety	4.24	0.60	84.80%	0.14
					12. Crime Prevention and Safety	4.20	0.65	84.00%	0.15
					13. Safety of Winter Activities	4.40	0.58	88.00%	0.13
Campus Transportation	4.23	0.88	85.20%	0.21	14. Pedestrian Transportation System	4.40	0.58	88.80%	0.13
					15. Diversity of Transportation Modes	4.33	0.57	86.60%	0.13

).

However, several indicators exhibited conceptual overlap. Three experts pointed out that “Winter Resource Integration and Utilization” overlapped with “Ice and Snow Resource Utilization”; four experts suggested that “Microclimate Regulation for Outdoor Activities” intersected with “Extreme Climate Protection Design” and “Spatial Scale”; and five experts considered that “Pedestrian Facility Quality,” “Pathway Obstacle Avoidance,” and “Winter Activity Safety” duplicated the meaning of “Pedestrian Traffic Safety.” To improve the distinctiveness and logical clarity of the indicators, this study removed the primary indicator “Characteristics of Campus Facilities in Cold Regions,” along with the following secondary indicators: “Extreme Climate Protection Design,” “Microclimate Regulation for Outdoor Activities,” “Pedestrian Facility Quality,” “Winter Activity Safety,” “Pathway Obstacle Avoidance,” “Pedestrian Pathway Interface Design,” and “Ice and Snow Resource Utilization.”

The second round of the Delphi consultation aimed to assign weights to each dimension.

Table 7. Weight Statistics from the Second Round of Expert Consultation.

Environmental Factors	Mean	Standard Deviation	Coefficient of Variation	Secondary Indicators
Climate Adaptability	19.5	3.45	0.18	1. Extreme Climate Environment Protection Design
Architectural Forms	26	4.12	0.20	2. Layout Design
				3. Spatial Scale
Natural Environment and Green Spaces	16	4.74	0.17	4. Seasonal Vegetation
				5. Utilization of Ice and Snow Resources
Campus Safety	16.5	5.87	0.21	6. Traffic Safety
				7. Crime Prevention and Safety
Campus Transportation	22	05.16	0.29	8. Pedestrian Transport System
				9. Diversity of Transportation Modes

presents the mean, standard deviation, and coefficient of variation of expert-assigned weights for each dimension in the second round of consultation. The results show that the mean value for Architectural Forms in Cold Regions was the highest (M = 26), indicating that experts generally regarded architectural form as the key factor influencing student health. This was followed by Campus Transportation in Cold Regions (M = 22), Climate Adaptability in Cold Regions (M = 19.5), Campus Safety in Cold Regions (M = 16.5), and Natural Environment and Green Spaces in Cold Regions (M = 16). The coefficient of variation for all primary indicators was below 0.30, suggesting a high level of consistency among expert opinions.

Through two rounds of Delphi consultations, this study ultimately developed an environmental health impact evaluation system consisting of six primary indicators and eight secondary indicators. The results highlight the central importance of architectural form, transportation safety, and climate adaptability in health-oriented campus design for cold regions, providing a reliable theoretical foundation for subsequent empirical research.

3.3. Case Study on Health Promotion Mechanisms in Cold-Region University Campuses

3.3.1. Case Selection and Research Methods

To examine how campus environmental improvements promote student physical activity and health under cold climatic conditions, this study selected four representative universities in northern China based on the principles of typicality and comparability as case study subjects, including Harbin Engineering University, Shenyang Jianzhu University, Northeast Forestry University, and Xi’an Eurasia University.

These institutions are located in cold or severely cold regions and have implemented various innovative strategies in campus spatial planning and environmental design to respond to climatic challenges, making them representative cases. The study summarizes and compares spatial strategies, behavioral responses, and health effects across four dimensions: climate condition utilization, built environment design, landscape and ecological strategy, and safety and accessibility design.

3.3.2. Analytical Framework

The analytical logic of this study is constructed around a “spatial intervention–behavioral motivation–health promotion” mechanism chain. It emphasizes how campus environmental design in cold regions can stimulate students’ physical activity and improve psychological well-being through the optimization of physical space. The corresponding relationships among the four analytical dimensions are shown in

Table 8. Dimensions, Intervention Types, Behavioral Responses, Health Promotion Mechanisms, and Case Examples.

Dimension	Type of Intervention	Behavioral Response	Health Promotion Mechanism	Case Example
Climate Condition Utilization	Climate resource transformation	Increased participation in outdoor activities	Enhanced physical activity level	Harbin Engineering University—Winter Carnival
Built Environment Design	Microclimate optimization and integrated circulation	Improved travel willingness and comfort	Promotion of daily walking	Shenyang Jianzhu University—Building layout; Xi’an Eurasia University—Wind-proof design; Harbin Institute of Technology—Corridor system
Landscape and Ecological Strategy	Aesthetic and nature-based experience	Enhanced emotional stability and pro-environmental behavior	Improved mental health	Northeast Forestry University—Four-season landscape design
Safety and Accessibility	Protective and guidance-oriented design	Extended activity duration and spatial range	Reduced risk perception and anxiety	Heilongjiang Institute of Technology—Hierarchical lighting system; Northeast Agricultural University—Anti-slip permeable pavement

.

3.3.3. Case Study Results and Analysis

Climate Adaptability

Cold climatic conditions are often regarded as major constraints on students’ use of campus spaces and participation in outdoor activities. However, when climatic features are reinterpreted as design resources that can be utilized, a transformation “from limitation to motivation” can be achieved. The Winter Carnival at Harbin Engineering University is a representative example. The university replans previously underutilized lawns, squares, and other spaces during winter as venues for ice and snow activities, making sports such as skiing, cycling, and ice sculpture making integral components of the campus’s winter culture.

This design strategy, by reconfiguring spatial utilization patterns, stimulates students’ willingness to participate in outdoor activities, extends the use cycle of public spaces on campus, and subtly promotes students’ physical exercise and social interaction. The proactive utilization of climatic conditions demonstrates an innovative pathway through which cold-region universities can transform unfavorable climatic conditions into opportunities for health promotion (see Figure 5).

Built Environment Optimization

The low-density and low-rise organic layouts commonly found in universities located in cold regions are conducive to creating horizontal and vertical spatial variations. Four main types of building enclosures—parallel, courtyard, U-shaped, and square-shaped—form spatial organizations that can improve the microclimate. Low-rise enclosed layouts reduce wind corridors by decreasing the spacing between buildings, thereby mitigating wind speed, minimizing heat loss between buildings, and enhancing overall campus comfort [76]. At Shenyang Jianzhu University (Hunnan Campus), the enclosed building configuration effectively reduces wind speed and improves daylighting conditions, significantly increasing the perceived thermal comfort of outdoor spaces during winter (Figure 6). Similarly, Xi’an Eurasia University incorporates windbreak hedges and human-centered seating design, which extend students’ outdoor stay duration (Figure 7).

Due to dispersed spatial layouts in some campuses, students and staff are exposed to cold conditions during long commutes, affecting travel comfort and spatial experience [77]. To address this, Harbin Institute of Technology implemented an extensive corridor system, enabling seamless connections between buildings and ensuring the continuity of pedestrian activities (Figure 8). Collectively, these design strategies establish an integrated “cold protection–mobility–physical activity” spatial system that promotes students’ daily physical engagement within a comfortable and safe campus environment.

Landscape and Ecological Strategies

During winter, the withering of vegetation and monotony of colors often make outdoor campus environments less appealing, thereby reducing students’ frequency of outdoor activities [78]. Northeast Forestry University maintains environmental vitality throughout the winter through four-season landscape design. The configuration of cold-resistant trees and perennial plants not only sustains ecological stability but also creates visually layered landscape compositions (Figure 9). The diversity of winter colors and plant forms provides positive psychological cues for students, alleviating winter fatigue and depressive moods while enhancing their motivation for outdoor activities. This strategy demonstrates that landscape is not merely an aesthetic decoration but also an important medium for stimulating students’ willingness to engage in physical activity and promoting psychological well-being.

Safety and Accessibility Enhancement

In cold regions, long winter nights and short daylight hours make campus lighting a critical element of design, not only to meet students’ basic activity needs but also to effectively prevent the formation of dark zones that may pose potential safety hazards [79]. The safety design of cold-region campuses directly influences the scope and duration of student activities. Heilongjiang Institute of Technology has improved nighttime visibility through layered lighting and color temperature control (Figure 10). Given the prevalence of snow and ice during winter, slipping accidents occur frequently, making the anti-slip performance of paving materials particularly important [80]. China Agricultural University employs anti-slip permeable pavement to reduce the risk of icing (Figure 11).

These improvements reduce the probability of potential accidents, enabling students to travel, exercise, and socialize safely during winter. The creation of a safe environment not only provides physical protection but also psychologically enhances students’ confidence in outdoor activities, becoming an important component of the health promotion mechanism.

3.3.4. Framework Validation and Theoretical Convergence

This study constructed a multi-stage research framework centered on LDA topic modeling, factor analysis, Delphi expert consultation, and case validation, forming a logical closed loop in which theoretical generation, data validation, expert refinement, and empirical feedback mutually reinforce one another. Through the LDA model, topic mining was conducted on literature related to university campuses in cold regions, identifying six major topics: climate, architectural form, campus facilities, natural environment and green spaces, safety, and transportation. This revealed the multidimensional structure through which campus environments influence student health and provided the theoretical foundation for subsequent quantitative analysis and expert consultation.

Subsequently, factor analysis was employed to verify the reliability and validity of the framework. Exploratory factor analysis (EFA) extracted six major factors with a cumulative variance contribution rate of 93.921%, confirming the framework’s internal consistency and structural soundness. However, several indicators exhibited lower factor loadings, suggesting that some factors possess regional or marginal characteristics specific to cold regions. This provided empirical evidence for refining the topic structure identified during the LDA stage.

Building on these findings, the Delphi method was used to optimize the framework through two rounds of expert consultation, consolidating and simplifying conceptually overlapping indicators. For example, “Microclimate Regulation for Outdoor Activities” was merged with “Protective Design for Extreme Climatic Conditions,” thereby strengthening the logical hierarchy and structural clarity of the indicator system. Finally, case studies of representative cold-region universities were conducted to empirically validate the optimized framework. The results demonstrated that the framework effectively captures the multidimensional interactions between campus environments and student health, exhibiting strong applicability and explanatory power.

In summary, the study established a closed-loop pathway from text mining to data validation to expert consensus to empirical reflection, constructing a dynamic mechanism of bidirectional feedback between theory and practice. This provides systematic theoretical support and methodological reference for research on campus environment optimization and health promotion in cold-region universities.

4. Conclusions and Recommendations

4.1. Empirical Findings on Environmental Intervention Strategies in Cold-Region Campuses

Grounded in the socio-ecological perspective, this study constructed a three-dimensional analytical framework of “built environment–climatic environment–perceived environment” to explore how environmental interventions in cold-region campuses can enhance the use of outdoor spaces during winter and promote students’ physical activity and health levels. The findings show that cold-region universities can optimize their winter campus environments through four key strategies: (1) activating seasonal spaces with ice and snow activities; (2) regulating microclimates through building layout; (3) enhancing visual attraction through four-season landscape design; and (4) improving safety through lighting and pavement design. These strategies collectively improve spatial accessibility, comfort, and attractiveness, stimulating students’ willingness to continuously engage in outdoor activities during cold seasons, thereby contributing to increased daily physical activity levels and serving campus health promotion goals.

4.1.1. Positive Utilization of Climatic Conditions

Most existing studies regard cold climates as obstacles to campus space use, advocating enclosed corridors or sheltered pathways to minimize exposure time and improve travel efficiency and thermal comfort [81]. However, through the case of Harbin Engineering University, this study found that cold climates themselves can serve as resources for spatial activation and cultural expression. Through activities such as winter carnivals and ice sculpture competitions, previously underused spaces are transformed into participatory and symbolic seasonal venues. This finding aligns with [82] notion that public spaces should encourage interaction and stay, and responds to [83] pathway of “context-driven physical activity interventions.”

Therefore, winter campus design should shift from passive cold protection to active utilization by introducing culturally distinctive, climate-responsive activities that create external environments promoting social engagement and physical activity, thereby extending the spatial use logic of cold-region universities under seasonal transitions.

4.1.2. Microclimate Regulation Through Built Environment

Adjusting building orientation can significantly improve winter solar access, while utilizing low-density enclosed layouts helps reduce wind speed and mitigate heat loss [76,84]. The case of Shenyang Jianzhu University further demonstrates that a composite spatial organization strategy centered on “enclosure + rotation + scale control” can achieve integrated regulation of heat, light, and airflow during the cold season, while maintaining ventilation and ecological corridor functions in summer—thus forming an all-season adaptive spatial system.

Compared with previous studies focusing on single-dimensional thermal comfort improvement [85], this study emphasizes the microclimatic regulatory capacity of building spatial systems at the cluster scale and regards it as crucial infrastructure influencing students’ outdoor activity behavior. This strategy facilitates the creation of “livable outdoor spaces” under extreme climatic conditions, thereby supporting more stable patterns of activity behavior.

4.1.3. Influence of Winter Landscape on Perceived Environment and Behavior

Campuses in cold regions commonly suffer from withered vegetation and monotonous colors during winter, leading to declining spatial attractiveness and negative psychological states among students [78]. Through analyzing the landscape practice of Jinqiao Garden at Northeast Forestry University, this study proposes a landscape strategy centered on "native hardy species + multilayer landscape structure + four-season color composition," creating a visually rich, structurally diverse, and atmospherically friendly ecological environment.

Compared with previous research emphasizing primarily botanical configuration or aesthetic presentation, this study further emphasizes the behavioral guidance role of perceived environments—particularly the positive influence of color dynamics and seasonal changes on emotion regulation and outdoor activity intention. This aligns with the research of [12] on restorative experiences and preference frameworks provided by natural landscapes, introducing a psychological response dimension into cold-region landscape design and reinforcing its indirect supportive role in health promotion.

4.1.4. Integrating Safety and Cultural Perception in Design

In cold regions, safety perception is closely associated with campus lighting and winter ground slip prevention. One study points out that campus outdoor lighting that balances functionality and aesthetics helps improve nighttime walking safety perception [86]. Additionally, research has examined the performance of anti-slip materials under cold climatic conditions, emphasizing their critical role in enhancing campus winter travel safety on ice and snow [87].

Research demonstrates that differentiated deployment of anti-slip materials (such as PC bricks and flamed stone surfaces) and precise placement of lighting systems in lawns, sports fields, and guiding pathways effectively improve perceived spatial safety. This strategy achieves a transition from “basic safety” to “experiential safety,” responding to [88] health city research framework on the relationship between built environment and walkability, facilitating students’ extension of physical activity behavior across broader spatial and temporal scales.

Overall, grounded in the three-dimensional socio-ecological environmental structure, this study emphasizes that through spatial organization of the built environment, functional activation of climatic resources, and psychological guidance of perceived environments, a composite outdoor environment system characterized by “usability–safety–attractiveness” can be constructed during cold seasons. These strategies not only enhance the utilization efficiency of winter campus spaces in cold-region universities but, more importantly, effectively stimulate students’ physical activity behavior through microclimate improvement, enhanced perceptual experience, and cultural belonging, promoting their transformation from static space use to dynamic health behaviors and providing sustainable environmental support pathways for campus health promotion.

4.2. Theoretical and Practical Contributions

Grounded in the socio-ecological perspective, this study established a three-dimensional interactive model of “built environment–climatic environment–perceived environment.” It breaks through the traditional research framework that treats climate merely as an external constraint and instead emphasizes the bidirectional feedback mechanisms between environmental systems and individual behaviors. This model not only reveals the multidimensional structure of environmental health effects in cold-region universities but also provides an operational analytical pathway for health-oriented campus planning research.

At the practical level, the study proposes four spatial intervention strategies: climate resource utilization, building layout optimization, four-season landscape design, and safety enhancement. These strategies demonstrate strong replicability and are particularly applicable to university campuses in Northeast China, Inner Mongolia, Xinjiang, and high-altitude regions. The concepts of “seasonal spatial transformation” and “functional hybrid configuration” emphasized in this research offer methodological support for developing climate-responsive campuses with strong adaptability and rich experiential qualities.

4.3. Research Limitations and Future Directions

Although this study proposes a multi-factor framework with both theoretical and practical value, certain limitations should be acknowledged. First, the research primarily relies on literature analysis and case-based generalization, lacking direct support from field observations and behavioral data. Second, the research subjects are concentrated in universities in cold regions of China and have not fully considered the influences of different cultural backgrounds and policy environments on the results. Third, the mechanisms through which architectural and landscape elements contribute to microclimate regulation in cold regions still lack quantitative verification based on precise climatic simulations. Overall, this research is conceptual exploration rather than experimental research, aiming to propose and preliminarily validate the theoretical logic of the three-dimensional framework—“built environment–climatic environment–perceived environment”—to lay a foundation for subsequent empirical and experimental studies.

Future research can strengthen user-centered empirical studies to quantify students’ perceptions and behavioral responses to various environmental intervention strategies. By integrating questionnaire surveys, behavioral tracking, and spatial behavior modeling, more robust evidence can be obtained to reveal the linkage mechanisms between environmental design features and health outcomes. Such studies will help deepen understanding of how specific spatial interventions affect user experiences and well-being, providing a more solid empirical basis for health-oriented campus design in cold-region universities. Additionally, advanced microclimate simulation tools such as ENVI-met and CFD can be employed to analyze how building layouts, vegetation configurations, and spatial form variations influence thermal comfort and microclimate regulation. Combined with intelligent environmental technologies such as adaptive lighting, snow-melting systems, and real-time environmental monitoring, future research can further promote the development of “smart cold-region campuses” with dynamic environmental responsiveness and health-supportive functions.

4.4. Conclusions

In summary, the design and management of winter campus spaces in cold-region universities should not be limited to passive responses to adverse climatic conditions but should achieve proactive adaptation and innovative creation in spatial layout, ecological construction, and cultural expression. Based on the systematic perspective of social ecology, this study constructed a three-dimensional analytical framework of “built environment–climatic environment–perceived environment,” revealing the synergistic mechanisms among multidimensional environmental elements. The research demonstrates that coldness is not the terminator of campus spatial vitality but can instead serve as a catalyst for activating place culture and environmental creativity. Through scientific planning and environmental guidance, cold-region universities have the potential to realize the functional transformation and emotional activation of “ice-and-snow campuses,” moving toward spatial development pathways with stronger adaptability, greater ecological friendliness, and more distinctive cultural characteristics.

It should be noted that the analytical framework proposed in this study remains at the theoretical exploration stage, primarily constructed based on literature materials and typical case analysis, and has yet to be validated through systematic field investigations or post-occupancy evaluations. Therefore, its applicability and effectiveness in practical operations still require further verification.

Future research can focus on empirical testing of this framework, integrating behavioral observation, questionnaire surveys, and spatial behavior modeling to explore its adaptability and predictive capacity across different climatic types, cultural backgrounds, and university types, thereby further refining the health-oriented spatial intervention system for cold-region universities and providing more universal and scientific theoretical support for subsequent design practices.