Impact of extreme temperature on hospitalization for cardiovascular diseases in Lanzhou, China

Abstract

Objectives

The purpose of this study was to examine the relationship between extreme temperature and the risk of hospitalization for cardiovascular diseases (CVDs).

Methods

A distributed lag nonlinear model (DLNM) in combination with a quasi-Poisson regression model was employed to assess the relationship between extreme temperature and risk of hospitalization for CVDs. This approach can be utilized to deal with lag and nonlinear effects. By comprehensively leveraging data information, it can explain the influencing factors from multiple aspects. Due to the complexity of the model, parameter estimation and model fitting typically require significant computational resources and extended processing time. Additionally, we identified the sensitive populations through subgroup analyses based on age and sex.

Result

Extremely low temperature (≤-10℃) (Relative Risk (RR) = 1.156, with a 95% confidence interval (CI) of 1.095–1.221), moderately low temperature (> 10℃, ≤-2℃) (RR = 1.132 95% CI: 1.091–1.174), moderately high temperature (≥ 20℃, < 28℃) (RR = 1.061 95% CI 1.039–1.084) and extremely high temperature (≥ 28℃) (RR = 1.124 95% CI: 1.080–1.169) were all associated with the increased risk of hospitalization for CVDs in the total population analysis, and low temperatures have a stronger effect than high temperatures. In the subgroup analysis, extremely low temperatures appeared to have a greater impact on females (RR = 1.280 95% CI:1.171-1.400) and < 65 age group (RR = 1.238 95% CI:1.147–1.339). Under extremely high temperature conditions, those more affected were males (RR = 1.130 95% CI: 1.073–1.190) and < 65 age group (RR = 1.222 95% CI: 1.163–1.285).

Conclusion

Low and high temperatures lead to an increased risk of hospitalization for CVDs with a lagged effect. Subgroup analyses indicated that females and < 65 age group were more sensitive to low temperatures, whereas males and < 65 age group were more sensitive to high temperatures.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-025-25804-4.

Introduction

Cardiovascular diseases (CVDs) are a category of serious health problems, encompassing coronary heart disease, stroke and hypertension and so forth. Currently, CVDs remain the leading cause of mortality and morbidity worldwide [1, 2]. Data showed that in 2019 alone, approximately 18.5 million people died from CVDs, constituting one-third of the global deaths in that year. Meanwhile, China has the largest number of CVD-related deaths in the world [3].

Since the Industrial Revolution, human activities have increased and greenhouse gases have been emitted in substantial quantities. Over the past century, global surface temperatures have increased by 0.7–1.4 degrees Fahrenheit and global sea levels have risen by 4–10 inches (10–25 centimeters) due to the melting of polar ice caps. The rise in global temperatures and sea levels has contributed to more frequent extreme weather events, such as heatwaves and cold snaps [4].

Many studies have confirmed that extreme temperature (e.g., cold snaps and heat waves) have a significant impact on health, and a large number of foreign studies have reported a relationship between extreme temperature and CVDs. A study focusing on 3,328 in-hospital cases of acute myocardial infarction in the north-central coastal region of Vietnam from 2008 to 2015 demonstrated that moderately low temperatures elevated the risk of hospitalization for myocardial infarction by 11% (with a 95% confidence interval (CI) of 0.91–1.35), whereas extremely low temperatures increased the risk by 25% (95% CI: 1.02–1.55) [5]. Meanwhile, a study carried out in Cyprus indicated that the positive correlation between high temperature and the mortality of CVDs was most conspicuous on the day of extremely high temperature occurred. Moreover, the relative risk (RR) remained significantly elevated for approximately one day after the occurrence of extremely high temperature [6]. The findings of numerous studies have also demonstrated that extreme temperatures affect the elderly and females to a greater extent [7–9].

The relationship between extreme temperature and many diseases has been widely acknowledged [10, 11]. However, similar studies conducted in different regions are region-specific due to climatic and physiological differences. This suggests that it is not feasible to directly extrapolate the temperature-disease relationship reported in one region to others without introducing potential errors [6].

Most studies have been conducted in developed countries, with research in our own country primarily focused on the eastern coastal regions, leaving the northwestern regions underexplored. Thus, we chose Lanzhou City, the capital of Gansu Province in northwestern China, as our study site. The objective of this study was to explore the relationship between temperature and CVDs hospitalization by utilizing the data on CVDs hospitalizations in 11 tertiary hospitals in Lanzhou City from 2013 to 2019. We included ischemic heart disease, heart failure, and hypertension in our study. Ischemic heart disease refers to a condition caused by myocardial ischemia, hypoxia, or necrosis due to factors such as coronary atherosclerosis. Heart failure is a syndrome characterized by structural or functional abnormalities of the heart, leading to inadequate cardiac output to meet the body’s metabolic needs, often resulting in pulmonary and/or systemic congestion. Hypertension is a clinical syndrome primarily characterized by elevated systemic arterial blood pressure, which may be accompanied by functional or structural damage to organs such as the heart, brain, and kidneys.

We also analyzed the study population by age and sex, which helps to identify high-risk groups and take more effective preventive measures. Furthermore, the results of this study may provide a basis for the formulation of appropriate protective policies and contribute to the evaluation of potential pathophysiologic mechanisms underlying the association between extreme temperature and CVDs.

Methods

Study region

Lanzhou, the capital of Gansu Province, is the political, economic, cultural, and transportation hub of the region. It holds a prominent position and exerts significant influence in Northwest China. Additionally, Lanzhou is located at the geometric center of the People’s Republic of China, at the junction of the Tibetan Plateau and the Loess Plateau, making it a region of unique geographical significance. Here, the annual temperature range and daily temperature range are large, and it is less affected by the regulation of the ocean. It is relatively cold in winter, with dry northwest winds blowing; while it is relatively warm in summer, with moist southeast winds blowing. The annual precipitation is between 200 and 400 mm and concentrates in summer. Therefore, it has a typical temperate semi-arid continental monsoon climate. The average annual temperature is 10 ℃ and the average annual precipitation is 360 mm.

Data collection

Demographic data

The data of this study were sourced from the hospitalization information of local residents who were diagnosed with CVDs in 11 tertiary hospitals in the urban area of Lanzhou City from January 1, 2013 to December 31, 2019, including information such as sex, date of birth, current address, time of hospitalization, time of discharge, inpatient diagnosis, and disease code. We included the data on CVDs in our study according to the International Classification of Diseases, 10th edition (ICD-10). The CVDs included in this study are as follows: ischemic heart disease (ICD-10: I20-I25), heart failure (ICD-10: I50.0-I50.1; I50.9), and hypertension (ICD-10: I10-I15). We conducted rigorous data cleaning and quality control procedures using SPSS (version 26.0) to establish a final analytic cohort of confirmed Lanzhou residents with complete demographic documentation. The standardized protocol included: (1) duplicate record detection through exact matching of patient identifiers (patient ID, admission date, and ICD-10 diagnosis codes) using the ‘Identify Duplicate Cases’ algorithm; (2) non-local residents were systematically excluded based on current residential address verification; (3) multiple imputation (m = 5) for missing sex (categorical) and age (continuous) data using predictive mean matching and logistic regression, respectively; and (4) preservation of original ICD-10 codes given minimal missingness (< 1% of cases).

Meteorological data

The meteorological data are sourced from Lanzhou Municipal Meteorological Bureau. The specific data include daily average temperature (℃) and daily average relative humidity (RH) (%). Meteorological data are collected using professional meteorological observation instruments certified by China Metrology, and professional meteorological observers are responsible for data quality audit and control to ensure the completeness and accuracy of the observation data. When missing data or outliers are encountered, the stl () function is first used to perform a robust regression decomposition of the data and identify outliers, and after manual review and confirmation, the outliers are deleted and the average value is used to fill in the missing values.

Air pollutant data

The air pollutant data were obtained from the four air quality monitoring stations (the station at the Institute of Biological Products, the station at the Railway Design Institute, the station at the Workers’ Hospital and the station at the Lanlian Hotel) set up by the State Environmental Monitoring General Station in Lanzhou City. The monitoring data are the continuous hourly concentrations of PM₁₀, NO₂ and SO₂ over a 24-hour period. The concentration of pollutants was measured in accordance with the “Ambient Air Quality Standards” (GB 3095 − 2012) issued by the National Environmental Protection Agency. Pollutant measurements were conducted using standardized techniques: gravimetric analysis for PM₁₀, formaldehyde absorption coupled with para-rosaniline spectrophotometry and mercuric chloride salt absorption with para-rosaniline spectrophotometry for SO₂, along with hydrochloric acid naphthylethylenediamine spectrophotometry for NO₂. We calculated the daily average concentration of each pollutant at each monitoring station based on the monitoring data, and then took the average of the data from the four monitoring stations as the daily average concentration of each air pollutant. All ambient air pollutants were measured in accordance with China’s national quality control standards.

Statistical analysis

Given established evidence of nonlinear and delayed associations between temperature and cardiovascular disease (CVD) morbidity (Reference), we employed a Distributed Lag Nonlinear Model (DLNM) to quantify temperature-CVD hospitalization relationships. Given established evidence that temperature impacts cardiovascular diseases in a nonlinear manner with lag effects [12, 13], we adopted the Distributed Lag Nonlinear Model (DLNM) to analyze the relationship between temperature and the hospitalization situation of CVDs. Meanwhile, the hospitalization events of CVDs are low-probability events, which are align with the occurrence of Poisson distribution. However, preliminary analyses revealed overdispersion in the data, as the observed variance exceeded the variance expected under the Poisson distribution. To address this, we applied the quasi-Poisson regression model, an extension of the Poisson regression that effectively handles overdispersion. Compared to other approaches like method negative binomial regression, the quasi-Poisson model provides a straightforward adjustment for overdispersion without requiring additional distributional assumptions. It can better fit the distribution characteristics of the actual data and improve the accuracy and reliability of the model. By combining the DLNM with the quasi-Poisson regression model, we conducted a time-series analysis tailored to the unique characteristics of our data [14, 15].

To control for confounding factors, we included some air pollutants and relative humidity (RH) in the model. Since these factors may exhibit covariance, we conducted a Spearman correlation analysis. According to previous articles [16], when the correlation coefficient between two factors is exceeds 0.8, they should not be included in the same model. At this point, we will screen the factors. If two variables are highly correlated, and one of the variables is theoretically or contextually less important for the research purposes, we may consider removing that variable from the model. However, when two or more variables are highly correlated and both hold significant theoretical relevance, we would consider combining them into a composite variable.

The specific model is as follows:

log[E(Y_t)]= α + βTemp_t,l+ ns(meteorological factors_t,dƒ) + ns(dare_t,dƒ) + ns(air pollutant factors_t,dƒ)+ DOW_t+ Holiday_t

Where, t denotes the date of observation during the study period, Y_t is the expected number of CVDs related hospitalization on day t; α is the intercept; β is the coefficient vector of the cross basis; ns is the natural cubic spline function; l is the hysteresis days; Temp_{t, l} is a two-dimensional cross-basis matrix generated by DLNM for fitting nonlinear temperature effects and hysteresis relations; rh_t indicates the average daily relative humidity on day t; PM_{10 t}, SO_{2 t}, and NO_{2 t} represent the concentrations of PM₁₀, SO₂ and NO₂ on day t, respectively; DOW_t and Holiday_t are dummy variables for “day of the week” and “holiday”, respectively, to adjust for differences in daily baseline hospitalization rates. In our study, we choose the degree of freedom (df) based on the minimum value of Quasi Akaike Information Criterion (Q-AIC), and a smaller value of Q-AIC indicates a better model.

We chose 7 dfs for the time variable in the model per year, and 3 dfs for relative humidity, and contaminants (PM₁₀, NO₂ and SO₂) [17]. Based on previous research and literature [18], we set the maximum lag to 21 days to fully assess the overall temperature effect.

To define the range of extreme temperatures, we included the daily average temperature in the basic model. By fitting the exposure-response relationship curve between temperature and the number of daily hospitalizations for CVDs, we determined the appropriate temperature and used it as the reference temperature. The 5th percentile (−10℃), 25th percentile (−2℃), 75th percentile (20℃), and 95th percentile (28℃) of the daily average temperature were taken as the extremely low temperature (≤−10℃), moderately low temperature (>−10℃, ≤−2℃), moderately high temperature (≥ 20℃, < 28℃), and extremely high temperature (≥ 28℃) respectively.

Subsequently, in order to identify the susceptible population, we also conducted a stratified analysis of this population by sex (male, female) and age (< 65 age group, ≥ 65 age group).

Sensitivity analysis

To assess the stability of the model, we conducted a sensitivity analysis. This involved systematically varying the degrees of freedom for time trends, air pollutants, and meteorological factors, as well as the maximum lag days, to examine the model’s response to these changes. If the model shows high sensitivity to certain parameters, it may indicate that the model is unstable or overfitted. We took into account previous studies [19–21], the model’s fit, and the characteristics of the current data.

Results

Descriptive summary

Between 2013 and 2019, 268,208 hospitalized patients due to CVDs in 11 tertiary hospitals in Lanzhou City were included in the study, of which 157,383 were males and 110,825 were females. The number of people < 65 years old was 150,183, and there were 118,025 people ≥ 65 years old. Meanwhile, Table 1 shows that there were 105 hospitalizations in an average day, and the daily mean values of PM₁₀, SO₂, and NO₂ were 119.37 µg/m³, 22.54 µg/m³, and 45.51 µg/m³, respectively. The average temperature and humidity were 11.39 °C and 50.63%, respectively. The maximum temperature was 30.4℃ and the minimum temperature was − 12.3℃.

Table 1.

Descriptive statistics of inpatients with CVD, air pollutants and meteorological factors in Lanzhou during 2013–2019

Variables		Total	Mean	SD	Min	Median	Max
Demographic factors		268,208	105	68.47	0	101	419
Sex	Male	157,383 (58.7%)	62	40.53	0	58	254
	Female	110,825 (41.3%)	43	28.74	0	42	179
Age	< 65	150,183 (56.0%)	59	40.19	0	56	228
	≥ 65	118,025 (44.0%)	46	29.37	0	44	212
Air pollutants
PM10(µg/m3)		302830.55	22.54	14.97	0.00	44.31	146.60
SO2(µg/m3)		57190.49	45.51	17.91	0.00	18.42	115.51
NO2(µg/m3)		115465.94	119.37	86.71	0.00	102.85	1484.54
Meteorological factors
Temperature(℃)		29104.25	11.39	9.84	−12.30	12.89	30.40
Rh(%)		129398.74	50.63	15.31	11.71	51.00	96.09

SD Standard deviation, Min Minimum, Max Maximum, Rh Relative humidity

Table 2 shows the results of the spearman’s correlation analysis between the risk of hospitalization, meteorological factors, and air pollutants. The results showed that the correlations among all the factors were < 0.8, and there was no covariance. Therefore, all of them could be included in the model.

Table 2.

Spearman’s correlation between inpatients with CVD, air pollutants and meteorological factors in Lanzhou, Gansu, during 2013–2019

Variables	Case	Temperature	Rh	PM10	SO2	NO2
Case	1	0.004	0.012	−0.035	−0.105**	0.172**
Temperature	0.004	1	−0.013	−0.214	−0.61	−0.3
Rh	0.012	−0.013	1	−0.302**	−0.201**	−0.089**
PM10	−0.035	−0.214**	−0.302**	1	0.318**	0.176**
SO2	−0.105**	−0.610**	−0.201**	0.318**	1	0.434**
NO2	0.172**	−0.300**	−0.089**	0.176**	0.434**	1

^*P < 0.05 ^**P < 0.01

Figure 1 is a contour plot of temperature, lag days and risk of hospitalization. The graph shows a non-linear relationship between temperature and the risk of hospitalization at different lag days, with both high and low temperatures causing an increase in the risk of hospitalization, but the risk of hospitalization caused by low temperatures is lower and the lag time is shorter compared to high temperatures.

Fig. 1.

The contour plot of the association between temperature and CVDs hospitalizations

Figure 2 shows the exposure-response curves of temperature and the risk of hospitalization on the exposure day as well as at different cumulative lag days (Lag0-3, Lag0-7, Lag0-14, Lag0-21). Except for the situation of Lag0-21, the exposure-response curves all present a “U” shape, indicating that both low and high temperatures can contribute to an increase in the hospitalization risk of CVDs.

Fig. 2.

Exposure-response curves of temperature and CVDs hospitalizations risk in different lag days

Figure 3 shows that there was a cumulative lag effect in the risk of CVDs hospitalization due to either low, moderately low, moderately high, or high temperatures. Among them, the cumulative lag effect of high temperature lasts the longest. In the case of low temperature, the cumulative lag effect reaches its maximum RR value of 1.164 ((95% CI: 1.098–1.233) on Lag0-8. In the case of moderately low temperature, the cumulative lag effect reaches its maximum RR value of 1.138 (95% CI: 1.097–1.181) on Lag0-9. While in the cases of moderately high temperature and high temperature, the cumulative lag effects reach their highest values on Lag0-6 and Lag0-12 respectively, with the RR being 1.062 (95% CI: 1.039–1.085) and 1.120 (95% CI: 1.077–1.166). Specific details can be seen in the Supplement materials. (Table S1)

Fig. 3.

Exposure-response curves of lag days and CVDs hospitalizations risk in different temperature

Subgroup analysis

We grouped the hospitalized patients by age and sex to conduct subgroup analysis. Fig. 4 illustrates the specific situations of the sex and age subgroups. The low temperature environment had a greater impact on females and < 65 age group. Among them, the maximum RR values for low temperature, which were 1.285 (95% CI: 1.175–1.405) and 1.245 (95% CI: 1.153–1.344) respectively, both appeared on Lag0-8. Meanwhile, the maximum RR values for moderately low temperature were 1.199 (95% CI: 1.132–1.269) and 1.228 (95% CI: 1.168–1.291), which appeared on Lag0-9 and Lag0-10 respectively.

Fig. 4.

Exposure-response curves of lag days and CVDs hospitalizations risk of different age and sex in different temperature

In contrast, under high temperature conditions, males and < 65 age group were more affected. Among them, the maximum RR values for high temperature were 1.127 (95% CI: 1.071–1.187) and 1.216 (95% CI: 1.157–1.279), which appeared on Lag0-12 and Lag0-15 respectively. While the maximum RR values for moderately high temperature were 1.063 (95% CI: 1.034–1.093) and 1.068 (95% CI: 1.036–1.102), which appeared on Lag0-6 and Lag0-7 respectively.

Sensitivity analysis

To assess the robustness of the model, we adjusted the degrees of freedom for the time variables (df = 6, 7, 8), different pollutants and meteorological factors (df = 2, 3, 4), as well as the maximum lag days (max lag = 20, 21, 22), and the results of the study were unchanged, suggesting that the model calculations were good. The detail can be seen in Figure S1.

Discussion

Our DLNM analysis revealed a U-shaped temperature-CVD hospitalization association in Lanzhou (2013–2019), with extreme heat (RR = 1.42, 95% CI: 1.32–1.53) and cold (RR = 1.38, 95% CI: 1.28–1.49) showing greatest risks, consistent with national studies [7, 22].

This result is consistent with findings from several studies. Exposure to extreme heat may lead to dehydration, salt depletion, and increased surface circulation, which may result in the failure of thermoregulation [23]. Extreme heat may also be associated with elevated blood viscosity, cholesterol levels and sweating threshold [24]. Moreover, at high temperatures, blood flows from the core to the surface of the skin to cool the body, and cardiac output increases significantly as blood vessels dilate [25].

In this study, both extremely low temperatures and moderately low temperatures can also increase the risk of hospitalization for CVDs. The results of this study suggest that exposure to extremely low temperatures poses a higher risk of hospitalization for coronary heart disease compared to exposure to moderately low temperatures. Our findings are consistent with previous research, a study in the rural Northwest showed that the RR of extremely and moderately low temperatures for hospitalizations for CVDs were 1.916 (95% CI: 1.446–2.537) and 1.834 (95% CI: 1.449–2.322).

Several potential pathological mechanisms have been proposed. Recent research indicates that Cryoglobulins, a type of immunoglobulin that precipitates in cold conditions and redissolves at 37 °C, can induce vasculitis and occlusive vasculopathy when deposited on vascular endothelium under low-temperature and high-concentration conditions [26]; Exposure to extremely cold temperatures has been linked to increases in blood pressure, blood cholesterol, heart rate, plasma fibrinogen, platelet viscosity and peripheral vasoconstriction [27, 28]. All of these factors contribute to an increased risk of hospitalization for CVDs. Additionally, a study has shown that cold exposure is associated with elevated low-density lipoprotein (LDL) cholesterol levels and reduced high-density lipoprotein (HDL) cholesterol levels, both of which are critical contributors to the heightened incidence of CVDs [29].

The observed lag patterns exhibit distinct geographical variations, with Hong Kong’s tropical monsoon climate inducing immediate heat-related health impacts (peak effect at Lag0-1) [30], while Lanzhou’s continental plateau climate demonstrates significantly delayed cardiovascular responses (peak at Lag0-12). This temporal discrepancy likely stems from the combined effects of thermal adaptation and behavioral regulation: the substantial difference in mean annual temperature between Lanzhou (11.4℃) and Hong Kong (23.4℃) promotes physiological cold adaptation that prolongs cardiovascular stress duration, compounded by Lanzhou’s municipal heating system which maintains stable indoor temperatures (18–22℃) during the 150-day winter period (November 15-March 15), thereby attenuating acute outdoor cold exposure effects.

We categorized hospitalized patients by age and sex, and the study results revealed that the risk of hospitalization for CVDs varied among individuals with different characteristics (sex and age) within the same temperature range.

Cold vulnerability was higher in females and < 65 years, potentially linked to centralized heating reducing elderly exposure. A study conducted in Beijing also supports this idea [31]. A study in Scotland confirms that females are at higher risk of myocardial ischemia, cardiac arrhythmias and elevated blood pressure in extremely cold temperatures [32]. Meanwhile, a physiologic study showed that females are more prone to Raynaud’s phenomenon, a strong cold-induced constriction of cutaneous arteries, under cold conditions. The study also found that estrogen increases the expression of smooth muscle α2C-adrenoceptors and enhances cold-induced constriction of cutaneous arteries, a condition more commonly observed in females than in males [33, 34].

The results show a higher risk of high temperatures in males and < 65 years, which is not quite consistent with some of the earlier articles [35–37]. We believe that some social factors are responsible for the discrepancy. Adult males are often exposed to high temperatures outdoors for significantly longer periods than the elderly and females due to their work, social interactions, and physical activities [38–40]. There are more males than females who engage in unhealthy habits such as smoking and drinking. Several surveys have shown that alcohol consumption is strongly associated with the development of CVDs [41, 42]. Meanwhile, the elderly spend significantly less time outdoors during heat warnings [43] And as the use of air conditioning becomes more widespread, the elderly are increasingly less affected by high temperatures. These factors may also play a significant role in contributing to the results of this study [44].A survey conducted across 12 Chinese cities also confirmed the idea that the effect of temperature diminishes with age [45].

There were several strengths in this study. Firstly, we selected patients hospitalized for CVDs in 11 tertiary hospitals in Lanzhou City from 2013 to 2019 as the study population, providing a large, stable and representative sample size. Secondly, we used DLNM to examine the relationship between temperature and CVDs, while adjusting for confounding factors such as relative humidity and air pollution. This approach enhanced the reliability of the results. Finally, we examined differences in effect sizes across age and sex populations. Therefore, this research can provide valuable insights for local disease prevention and control efforts.

There are some limitations in our study. Firstly, although air quality monitoring stations provide valuable data, it is important to recognize their limitations. The data from these monitoring points do not reflect individual exposures, and the exposure levels of individuals in certain occupations may differ significantly from those recorded at the monitoring stations. Additionally, it is challenging for only four monitoring stations to fully represent air quality across the entire city. Therefore, in future studies, we will increase the number of monitoring stations and optimize their distribution. Additionally, we will incorporate multiple statistical methods and models for calculations, rather than relying solely on the mean value approach, to mitigate the limitations associated with using a single method. Secondly, our study was conducted solely in Lanzhou City, and due to regional climatic and ethnographic differences, it is challenging to generalize the results to other areas. Lastly, we did not integrate the possible effects of other variables such as comorbidities, genetic factors, and personal habits on the results.

Conclusion

The results of this study indicate that both low and high temperatures cause an increased risk of hospitalization for CVDs with a lagged effect, while females and < 65 age group are more sensitive to low temperatures, meanwhile males and < 65 age group are more sensitive to high temperatures. Therefore, the results of our study can offer valuable insights for disease prevention departments and serve as a reference for relevant policy-making bodies.

Abbreviations

CVDs

Cardiovascular Diseases

DLNM

Distributed Lag Non-linear Models

RR

Relative risk

CI

Confidence Interval

RH

Relative Humidity

DOW

Day of the week

SD

Standard deviation

Min

Minimum

Max

Maximum

LDL

Low-density lipoprotein

HDL

High-density lipoprotein

Authors’ contributions

Conceptualization: Miaoxin Liu, Anning Zhu, Jingze Yu, Ye Ruan Data curation: Miaoxin Liu, Anning Zhu Formal analysis: Ke Xu, Bin Luo, Jingping Niu. Funding acquisition: Ye Ruan Methodology: Rentong Chen, Tong Liu, Li Zhang. Project Administration: Ye Ruan Software: Miaoxin Liu, Anning Zhu, Jingze Yu Supervision: Rentong Chen, Ye Ruan Validation: Ke Xu, Bin Luo, Jingping Niu. Visualization: Miaoxin Liu, Ke Xu. Writing – original draft: Miaoxin Liu, Anning Zhu, Jingze Yu Writing – review & editing: Ke Xu, Li Zhang, Ye Ruan. All authors read and approved the final manuscript.

Funding

This work was supported by the Fundamental Research Funds for the Central Universities (lzujbky-2020-9).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This study conducted in accordance with the Declaration of Helsinki, collected hospitalization data without identifiable personal information. The Institutional Review Board (IRB) of the School of Public Health, Lanzhou University, exempted the study from ethics approval and consent, considering the absence of identifiable participant information. The researchers ensured the privacy and confidentiality of the data throughout the study, adhering to the guidelines and regulations stipulated in the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.