Inferior metabolic control in isolated diabetic patients under the context of multimorbidity management: a comparative study

Abstract

Purpose

To explore the metabolic control status of patients with type 2 diabetes mellitus (T2DM) with or without hypertension or hyperlipidemia under the management of metabolic management centers (MMC).

Method

Eligible patients with T2DM from the Yan’an University Affiliated Hospital Branch of the MMC were retrospectively enrolled and categorized into non-hypertensive and hypertensive groups, as well as non-hyperlipidemic and hyperlipidemic groups. Propensity score matching (PSM) was used to balance the baseline confounding variables. Intergroup comparisons of metabolic indicators were conducted and correlation analyses were performed for these indicators.

Result

After PSM to balance baseline confounding factors, T2DM patients with hypertension or hyperlipidemia showed significantly better control of key metabolic indicators than those without these comorbidities. Specifically, the comorbid groups had lower levels of glycated hemoglobin (HbA1c), postprandial glucose, total cholesterol, and low-density lipoprotein cholesterol (LDL-c). Additionally, the insulin resistance index (HOMA-IR) correlated with triglyceride and high-density lipoprotein cholesterol (HDL-c) levels, further reflecting the metabolic advantages associated with comorbidities under standardized management.

Conclusion

T2DM patients without hypertension or hyperlipidemia have poorer metabolic control under MMC management, emphasizing the need for tailored management strategies for this “potentially underprioritized group.”

Clinical trial number

Not applicable.

Introduction

Diabetes, a chronic metabolic disorder, is spreading worldwide at an alarming rate. According to data from the World Health Organization (WHO), the number of people with diabetes has skyrocketed from 200 million in 1990 to 830 million in 2022, which is more than quadrupling in just over 30 years. The prevalence of diabetes among adults aged ≥ 18 years has surged from 7% in 1990 to 14% in 2022 [1, 2]. The number of adults aged 20–79 with diabetes worldwide will reach 589 million in 2025, and it is projected that the number of patients will increase to 853 million by 2050 [3]. This growth trend is particularly pronounced in low- and middle-income countries, where the prevalence is rising much faster than that in high-income countries. This global epidemiological trend has imposed substantial burdens on public health systems and hindered socioeconomic progress worldwide.

The harmful effects of diabetes are both extensive and serious. Prolonged hyperglycemia, such as a hidden time bomb, gradually erodes various organ systems in the human body. Diabetic retinopathy can cause severe visual impairment or even blindness in patients, and is one of the main causes of blindness in the working-age population [4]. If diabetic nephropathy is not effectively controlled, it will gradually progress to end-stage renal disease, greatly increasing the suffering of patients and the healthcare burden [5]. Therefore, the cardiovascular system is severely affected. The risk of developing cardiovascular diseases, such as coronary heart disease, stroke, and heart failure, is significantly increased in diabetic patients, making cardiovascular diseases the leading cause of death [6, 7]. In 2021, diabetes and kidney disease caused by diabetes resulted in more than 2 million deaths [8], and approximately 11% of cardiovascular deaths were attributed to hyperglycemia [7].

Among the diabetic patient population, comorbidities with other diseases are extremely common, greatly increasing the complexity of disease management [9, 10]. Many clinical studies have shown that chronic diseases, such as hypertension, dyslipidemia, obesity, and non-alcoholic fatty liver disease, often coexist with diabetes [9]. According to relevant statistics, approximately 60% of patients with type 2 diabetes mellitus (T2DM) have hypertension [11], and more than 60% have dyslipidemia [12]. These comorbid diseases do not exist in isolation, but are intertwined and interact with each other, jointly forming a complex “metabolic syndrome” network. For example, hypertension can further aggravate vascular damage in patients with diabetes and promote atherosclerosis. Dyslipidemia can interfere with the normal function of insulin, exacerbating insulin resistance and making glycemic control even more difficult.

Global management of diabetes is not optimistic [13–15]. Although modern medicine has made many advancements in diabetes treatment, such as the development of various hypoglycemic drugs, innovation in insulin treatment technology, and the promotion of lifestyle intervention concepts, achieving ideal disease control remains extremely difficult. Among the diagnosed patients, only a portion could control their blood sugar levels within the ideal range. In low-and middle-income countries, owing to the constraints of scarce medical resources, low health awareness among patients, and economic burdens, the treatment coverage and control rates of diabetes are even lower [16]. Even in high-income countries with relatively abundant medical resources, with the continuous increase in the prevalence of diabetes, especially the rapid growth of the young population, the healthcare system is under enormous pressure.

Currently, there are many unresolved controversies in the research field regarding the management and progression of diabetes with comorbid diseases. The efficacy and safety of different drugs vary significantly between patients with diabetes and multiple comorbidities. For example, in diabetic patients with cardiovascular disease, traditional hypoglycemic drugs, while controlling blood sugar, may have adverse effects on the cardiovascular system owing to risks such as hypoglycemia and weight gain [17, 18]. New hypoglycemic drugs, such as sodium-glucose cotransporter 2 (SGLT2) inhibitors [19] and glucagon-like peptide-1 (GLP-1) receptor [20] agonists, have shown some advantages in improving cardiovascular outcomes, but they are not suitable for all patients, and their long-term safety and cost-effectiveness still require further study [21]. In terms of lifestyle interventions, although diet control and regular exercise are recognized as the cornerstones of diabetes management, ensuring that patients adhere to them in the long term and obtain continuous health benefits remains an urgent problem. Some studies have pointed out that even if patients can follow lifestyle intervention recommendations in the short term and improve blood sugar and metabolic indicators in the long term, due to the difficulty and complexity of lifestyle changes, patient compliance often gradually decreases, greatly reducing the effectiveness of the intervention [22–24].

Against this complex and challenging background, this study aimed to analyze the control of metabolic indicators in patients with diabetes and different comorbid diseases. We hypothesized that compared with patients with type 2 diabetes without other comorbidities, patients with common comorbid diseases, such as hypertension and dyslipidemia, may receive more active and comprehensive treatment and management due to the clinical emphasis on comorbid diseases, and thus may perform better in the control of metabolic indicators. Through a detailed analysis of different subgroups of patients, this study aimed to reveal the internal relationship between the comorbid state of diseases and the control of metabolic indicators, provide a scientific basis for optimizing the clinical management strategy for diabetes and its comorbid diseases, and help improve the overall health of patients with diabetes.

Consistent with our hypothesis, under the standardized management of the Metabolic Management Center (MMC), patients with T2DM without hypertension or dyslipidemia demonstrated significantly worse control of key metabolic markers than those with these comorbidities. This difference persisted after balancing for baseline confounding factors, potentially due to the more active interventions received by comorbid patients.

Materials and methods

Subjects

This study retrospectively collected data from patients with metabolic disorders who were admitted to the Yan’an University Affiliated Hospital Branch of the MMC between January 2018 and December 2024. After rigorous screening based on the inclusion and exclusion criteria, 3,885 eligible adult patients with T2DM were enrolled as participants [2]. Based on the presence or absence of hypertension, the patients were categorized into non-hypertensive (Non-HP, n = 2,558) and hypertensive (HP, n = 1,327) groups. We further stratified these patients into a non-hyperlipidemic group (non-HL, n = 2,535) and a hyperlipidemic group (HL, n = 1,350) based on their hyperlipidemic status.

We adopted the following inclusion criteria: (1) Adults aged ≥ 18 years, in line with international guidelines for diabetes research, ensuring participants had fully developed metabolic regulatory mechanisms and stable disease phenotypes [2]; (2) Meeting the diagnostic criteria for T2DM formulated [2, 25], i.e., fasting blood glucose (FBG) ≥ 7.0 mmol/L, 2-hour blood glucose in oral glucose tolerance test (OGTT) ≥ 11.1 mmol/L, random blood glucose ≥ 11.1 mmol/L or glycosylated hemoglobin (HbA1c) ≥ 6.5%, and confirmed by clinical review; (3) Complete clinical data, including demographic information (age, gender, body mass index [BMI], education level, income), clinical characteristics (duration of diabetes, smoking status, drinking status), and laboratory test data (blood glucose, blood lipids, blood pressure, homeostasis model assessment of insulin resistance [HOMA-IR]), which can meet the data analysis needs of this study; (4) Duration of diabetes >12 months, and who had received standardized metabolic management20 (including regular follow-up, comprehensive assessments, and adjustments to intervention plans) at the MMC Yan’an University Affiliated Hospital branch for at least 6 months—this ensured uniformity in management protocols across all included patients. Subjects were excluded if: (1) patients aged less than 18 years or older than 75 years; (2) pregnant or lactating women; (3) patients in the acute phase of cardiovascular or cerebrovascular diseases; (4) acute complications of T2DM; (5) acute pancreatitis; (6) type 1 diabetes mellitus; and (7) sexually transmitted diseases, such as viral hepatitis, AIDS, and syphilis, as well as infectious diseases, such as tuberculosis, during the active stage.

This study adhered to the Declaration of Helsinki and international ethical norms for medical research. All clinical data of the research subjects are derived from the electronic database of the MMC Yan’an University Affiliated Hospital branch, and the collection and use of data have been approved by the Medical Ethics Committee of Yan’an University Affiliated Hospital (Ethics Approval No.: IIT-R-20250169). Due to the retrospective nature of the study, the informed consent form was waived with the approval of the Ethics Committee of Yan’an University Affiliated Hospital, and all patient data was anonymized to maintain confidentiality throughout the entire research process.

Clinical evaluation

All clinical data were recorded via the standardized data collection system of MMC Yan’an University Affiliated Hospital branch, and the detection procedures of each indicator strictly followed current international and domestic clinical testing standards.

Collection of demographic and clinical characteristics

Patient age, gender, education level (divided into ≤ 12 years/> 12 years), annual income (in thousands of Chinese yuan), and duration of diabetes (from diagnosis to inclusion in the study, accurate to months) were extracted from the electronic medical record system, and smoking status was classified into current smoking (≥ 7 cigarettes per week for > 6 months), occasional smoking (< 7 cigarettes per week or cumulative smoking <6 months), and non-smoking status, which was similarly classified into current drinking, occasional drinking, and non-drinking.

Anthropometric indicators

Height and weight measurement: Patients were measured without shoes and hats, wearing light clothing, using calibrated height and weight meters (accuracy of 0.1 cm and 0.1 kg, respectively). The body mass index (BMI) was calculated as weight (kg)/height (m²).

Blood pressure measurement: After the patient rested quietly for 5 min, the right brachial artery blood pressure was measured using a standardized electronic sphygmomanometer (regularly calibrated). Measurements were performed three times consecutively with an interval of 1–2 min between each measurement. The average of the last two readings was recorded as the systolic blood pressure (SBP) and diastolic blood pressure (DBP), and hypertension was defined as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg.

Laboratory indicator detection

Blood sample collection: All patients had 5mL of elbow venous blood collected early in the morning after fasting (no food for at least 8 h). Among them, 3mL was used for serum separation (centrifuged at 3000 rpm for 10 min, stored at 4 °C, and tested within 2 h), and 2mL was used for glycosylated hemoglobin (HbA1c) detection.

FBG was detected by the glucose oxidase method using an automatic biochemical analyzer, with a normal reference range of 3.9–6.1 mmol/L and a detection coefficient of variation (CV) < 2%. Venous blood was collected 2 h after the patient consumed a standard meal to detect postprandial blood glucose (containing 75 g of carbohydrates) or post 75gm of anhydrous glucose using the same detection method as FBG, and a normal reference range of <7.8 mmol/L. HbA1c was detected by high-performance liquid chromatography (HPLC) using a glycosylated hemoglobin analyzer, with a reference range of 4.0%-6.0% and a detection coefficient of variation (CV) < 1%. The results were interpreted in accordance with the Chinese Guidelines for the Prevention and Treatment of Type 2 Diabetes (2020 Edition) (target value < 7.0%).

Lipid indicators: Total cholesterol (TC) and triglycerides (TG) are detected by the enzymatic method, with normal reference ranges of <5.2 mmol/L and <1.7 mmol/L, respectively. Low-density lipoprotein cholesterol (LDL-c) and high-density lipoprotein cholesterol (HDL-c) are detected by a direct measurement method, with normal reference ranges of <3.4 mmol/L and >1.0 mmol/L, respectively. HOMA-IR was calculated using fasting insulin (FINS) and FBG levels, using the formula HOMA-IR = (FBG × FINS)/22.5. FINS was detected using the chemiluminescence method, with a normal reference value of <2.5 mIU/L, and higher values indicated more severe insulin resistance.

Statistical analysis

The SPSS (version 26.0 and R 4.2.0 software were used for clinical data processing and statistical analysis. The Kolmogorov-Smirnov test (K-S test) was used to verify the normality of the measurement data. Normally distributed continuous variables are presented as the mean ± standard deviation (±s). Between-group comparisons at baseline were performed using the independent samples t-test, whereas paired t-tests were used for comparisons after Propensity score matching (PSM). Non-normally distributed continuous variables are presented as median (interquartile range) and were compared using the Mann-Whitney U test at baseline, and the Wilcoxon signed-rank test was used for comparisons after PSM. Categorical variables were presented as counts (percentages) and compared using the chi-square test, as appropriate.

Propensity score matching (PSM), a key method to control selection bias in this study, works by integrating multiple baseline variables such as age, gender, education level, income, smoking status, drinking status, and duration of diabetes into a “propensity score” (i.e., the probability of an individual having hypertension or hyperlipidemia) through a logistic regression model [24, 26]. Then, according to a caliper width of 0.2 times the standard deviation of the logit of the propensity score, the exposure group (with hypertension/hyperlipidemia) was matched with the control group (without hypertension/hyperlipidemia) with similar baseline characteristics. This can effectively balance the differences in key variables between the two groups, making the comparability between groups close to that of randomized controlled trials, thereby reducing the interference of baseline confounding on the results and more accurately reflecting the association between comorbidity status and metabolic indicators.

Result

The overall design, main results, and key findings of this study are shown in Fig. 1.

Fig. 1.

Flowchart of this study. T2DM, type 2 diabetes mellitus; MMC, metabolic management center; HP, hypertensive group; non-HP, non-hypertensive group; HL, hyperlipidemia group; non-HL, non-hyperlipidemia group

Demographic and clinical characteristics of hypertensive subgroups

As shown in Tables 1, 3,885 participants were included, with 2,558 in the non-hypertensive (non-HP) group and 1,327 in the hypertensive (HP) group. Significant differences were observed in multiple indicators between the two groups. The HP group was older (P < 0.001) and had a longer disease duration (P < 0.001). In terms of comorbidities, the HP group showed a significantly higher prevalence of hyperlipidemia (P < 0.001) and a higher rate of lipid-lowering drug use (P < 0.001).

Table 1.

Demographic and clinical characteristics of hypertension subgroups

	Non-HP(n = 2558)	HP(n = 1327)	P
Age(years)	50.6 ± 10.5	56.44 ± 8.6	< 0.001
Gender(male/female)	1655/903	860/467	0.946
Duration(months)	84.9(43.1,134.8)	118.5(60.6,182.6.3)	< 0.001
Income(k CNY/year)	29.5(22.0,36.4)	29.3(21.7,36.4)	0.579
Education > 12years			0.177
Yes	1316(52.6)	668(50.8)
No	1212(47.4)	659(49.7)
Smoking status			0.053
Current-smoker(n,%)	436(17.0)	268(20.2)
Occasional-smoker(n,%)	655(25.6)	330(24.9)
Non-smoker(n,%)	1467(57.3)	729(54.9)
Drinking status			0.035
Current-drinker(n,%)	246(9.6)	149(11.2)
Occasional-drinker(n,%)	1056(41.3)	495(37.3)
Non-drinker(n,%)	1256(49.1)	683(51.5)
Hyperlipidemia			< 0.001
Yes (n, %)	668(26.1)	1350(51.4)
No (n, %)	1890(73.9)	645(48.6)
Lipid-lowering drug use			< 0.001
Yes (n, %)	211(8.2)	368(27.7)
No (n, %)	2347(91.8)	959(72.3)
SBP(mmHg)	123.3 ± 14.4	136.9 ± 17.2	< 0.001
DBP(mmHg)	76.5 ± 9.5	81.9 ± 11.1	< 0.001
BMI(kg/m2)	24.6 ± 3.4	25.8±0.3.4	< 0.001
FBG(mmol/L)	7.24(6.0,8.8)	7.06(6.0,8.4)	0.016
PBG(mmol/L)	13.8(11.0,16.8)	13.4(11.0,16.4)	0.048
HOMA-IR	2.3(1.2,4.2)	2.5(1.4,4.4)	0.013
HbA1c(%)	8.5(7.1,10.2)	8.0(7.0,9.6)	< 0.001
TG(mmol/L)	1.6(1.1,2.5)	1.6(1.2,2.5)	0.211
TC(mmol/L)	4.6 ± 1.1	4.4 ± 1.2	< 0.001
HDL-c(mmol/L)	1.0(0.8,1.2)	1.0(0.8,1.2)	0.325
LDL-c(mmol/L)	2.4(1.8,3.1)	2.2(1.6,2.9)	< 0.001

Note: HP, hypertensive group; Non-HP, non-hypertensive group; k CNY, thousands of Chinese yuan; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; FBG, fasting blood glucose; PBG, postprandial blood glucose; HOMA-IR, homeostasis model assessment of insulin resistance; TG, triglyceride; TC, total cholesterol; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol

Metabolically, the HP group had higher systolic blood pressure, diastolic blood pressure, and body mass index (all P < 0.001), along with a higher HOMA-IR (P = 0.013). However, the HP group had lower HbA1c (P < 0.001), fasting glucose (P = 0.016), postprandial glucose (P = 0.048), total cholesterol (P < 0.001), and LDL-c (P < 0.001).

There were no significant differences in sex distribution, annual income, education level, smoking status, triglycerides, and HDL-c between the two groups (all P ≥ 0.05), whereas drinking status showed a significant difference (P = 0.035), with a higher proportion of current drinkers in the HP group.

Analysis of hypertensive subgroups after PSM

After PSM, 1,751 participants were included in the non-hypertensive group (non-HP) and 1,165 in the hypertensive group (HP), with well-balanced baseline characteristics between the two groups.

Figure 2 illustrates the effectiveness of PSM in balancing baseline covariates between groups, as quantified by standardized mean differences (SMD). Before matching, the absolute SMD values of covariates such as age and disease duration were relatively large, indicating significant baseline differences. After matching, the absolute SMD values of all covariates (including age, sex, disease duration, income, education level, smoking, and drinking status) decreased to below 0.1, demonstrating that selection bias in baseline characteristics was effectively controlled, providing a reliable basis for subsequent comparisons of metabolic indicators.

Fig. 2.

Absolute standard mean differences of each covariate before and after propensity score matching between hypertensive subgroups

Figure 3 shows the distribution characteristics of propensity scores in both groups before and after PSM. Before matching, the propensity score distributions of the hypertensive and non-hypertensive groups had a low overlap, indicating inherent differences in the baseline characteristics between the two groups in the original cohort. After matching, the score distributions of the two groups overlapped highly with consistent ranges, confirming that PSM successfully constructed a research cohort with comparable baselines, ensuring that the subsequent metabolic differences observed were more likely to be directly related to hypertensive status.

Fig. 3.

Comparison of propensity scores before and after propensity score matching between hypertensive subgroups. HP: hypertensive group; Non-HP: non-hypertensive group

After balancing baseline characteristics, there were no significant differences between the two groups in age, gender distribution, disease duration, annual income, education level, smoking, or drinking status (all P ≥ 0.05).

However, the HP group still showed clinically significant differences in metabolic indicators, and the prevalence of hyperlipidemia was significantly higher than that in the non-HP group (P < 0.001), with a higher rate of lipid-lowering drug use (P < 0.001). Metabolically, the HP group maintained significantly higher systolic blood pressure (P < 0.001), diastolic blood pressure (P < 0.001), and BMI (P < 0.001), indicating unresolved risks related to hypertension and obesity. The HP group also had more severe insulin resistance, with a significantly higher HOMA-IR (P < 0.001). Notably, after PSM, the HP group showed significantly elevated triglyceride levels (P = 0.003), suggesting specific residual lipid metabolism abnormalities in T2DM patients with hypertension.

In contrast, the HP group achieved better glycemic control, with significantly lower HbA1c (P < 0.001), postprandial glucose (P = 0.009), total cholesterol (P = 0.007), and LDL-c (P < 0.001) (see Table 2).

Table 2.

Demographic and clinical characteristics of hypertension subgroups after PSM

	Non-HP(n = 1751)	HP(n = 1165)	P
Age(years)	54.9 ± 8.5	55.2 ± 8.3	0.638
Gender(male/female)	1172/579	769/396	0.604
Duration(months)	99.5(49.6,149.2)	101.3(53.7,170.3)	0.071
Income(k CNY/year)	29.5(22.1,36.5)	29.4(21.7,36.5)	0.705
Education > 12years			0.700
Yes (n, %)	895(51.1)	587(50.4)
No (n, %)	856(48.9)	578(49.6)
Smoking status			0.526
Current-smoker (n, %)	328(18.7)	228(19.6)
Occasional-smoker (n, %)	471 (26.9)	292(25.1)
Non-smoker (n, %)	952(54.4)	645(55.4)
Drinking status			0.317
Current-drinker (n, %)	179(10.2)	130(11.2)
Occasional-drinker (n, %)	710(40.5)	441(37.9)
Non-drinker (n, %)	862(49.2)	594(51.0)
Hyperlipidemia			< 0.001
Yes (n, %)	470(26.8)	620(53.2)
No (n, %)	1281(73.2)	545(46.8)
Lipid-lowering drug use			< 0.001
Yes (n, %)	168(9.6)	321(27.6)
No (n, %)	1583(90.4)	844(72.4)
SBP(mmHg)	124.1 ± 14.5	136.6 ± 16.9	< 0.001
DBP(mmHg)	76.4 ± 9.5	82.4 ± 11.0	< 0.001
BMI(kg/m2)	24.5 ± 3.3	25.9±0.3.4	< 0.001
FBG(mmol/L)	7.2(6.0,8.6)	7.1(6.1,8.4)	0.291
PBG(mmol/L)	13.9(11.2,16.8)	13.3(10.9,16.2)	0.009
HOMA-IR	2.2(1.2,4.0)	2.6(1.5,4.4)	< 0.001
HbA1c(%)	8.3(7.0,10.1)	7.9(6.8,9.5)	< 0.001
TG(mmol/L)	1.6(1.1,2.4)	1.7(1.2,2.5)	0.003
TC(mmol/L)	4.5 ± 1.1	4.3 ± 1.2	0.007
HDL-c(mmol/L)	1.0(0.8,1.2)	1.0(0.8,1.2)	0.065
LDL-c(mmol/L)	2.4(1.7,3.1)	2.2(1.6,3.0)	< 0.001

Note: PSM, propensity score matching; HP, hypertensive group; Non-HP, non-hypertensive group; k CNY, thousands of Chinese yuan; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; FBG, fasting blood glucose; PBG, postprandial blood glucose; HOMA-IR, homeostasis model assessment of insulin resistance; TG, triglyceride; TC, total cholesterol; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol

Demographic and clinical characteristics of hyperlipidemic subgroups

As shown in Table 3, the hyperlipidemic group was slightly older (P < 0.001) and had a higher annual income (P < 0.001) but a lower proportion of participants with education > 12 years (P < 0.001). The HL group also had a higher proportion of current smokers (P = 0.002) and drinkers (P = 0.019). Regarding comorbidities, the HL group had a significantly higher prevalence of hypertension (P < 0.001) and a higher rate of antihypertensive drug use (P < 0.001). Metabolically, the HL group had higher systolic blood pressure, diastolic blood pressure, and body mass index (all P < 0.001). It also showed higher HOMA-IR (P < 0.001), triglycerides (P < 0.001), and total cholesterol (P < 0.001), along with lower postprandial glucose (P < 0.001) and HbA1c (P < 0.001). There were no significant differences in sex distribution, disease duration, fasting glucose, or LDL-c levels between the two groups (all P ≥ 0.05), whereas HDL-c levels showed a significant difference (P = 0.009).

Table 3.

Demographic and clinical characteristics of hyperlipidemic subgroups

	Non-HL(n = 2535)	HL(n = 1350)	P
Age(years)	52.1 ± 10.6	53.4 ± 9.6	< 0.001
Gender(male/female)	1650/885	865/485	0.528
Duration(months)	99.6(48.8,158.3)	87.6(48.1,146.5)	0.094
Income(k CNY/year)	28.7(21.4,35.9)	30.7(23.0,37.3)	< 0.001
Education > 12years			< 0.001
Yes (n, %)	1395(55.0)	619(45.9)
No (n, %)	1140(45.0)	731(54.1)
Smoking status			0.002
Current-smoker (n, %)	420(16.6)	284(21.0)
Occasional-smoker (n, %)	662(26.1)	323(23.9)
Non-smoker (n, %)	1453(57.3)	743(55.0)
Drinking status			0.019
Current-drinker (n, %)	223(9.2)	162(12.0)
Occasional-drinker (n, %)	1015(40.0)	536(39.7)
Non-drinker (n, %)	1287(50.8)	652(48.3)
Hypertension			< 0.001
Yes (n, %)	645(25.4)	682(50.5)
No (n, %)	1890(74.6)	668(49.5)
Antihypertensive drug use			< 0.001
Yes (n, %)	497(19.6)	536(39.7)
No (n, %)	2038(80.4)	814(60.3)
SBP(mmHg)	126.5 ± 16.5	130.6 ± 16.7	< 0.001
DBP(mmHg)	77.5 ± 10.2	79.9 ± 10.5	< 0.001
BMI(kg/m2)	24.6 ± 3.4	25.8±0.3.3	< 0.001
FBG(mmol/L)	7.2(6.0,8.7)	7.2(6.1,8.5)	0.877
PBG(mmol/L)	14.0(11.3,17.0)	13.3(10.7,16.1)	< 0.001
HOMA-IR	2.2(1.1,4.1)	2.7(1.5,4.6)	< 0.001
HbA1c(%)	8.5(7.2,10.3)	8.0(6.9,9.6)	< 0.001
TG(mmol/L)	1.5(1.0,2.2)	2.0(1.4,3.0)	< 0.001
TC(mmol/L)	4.3 ± 1.1	4.6 ± 1.2	< 0.001
HDL-c(mmol/L)	1.0(0.8,1.2)	1.0(0.8,1.2)	0.009
LDL-c(mmol/L)	2.3(1.7,3.0)	2.4(1.8,3.1)	0.167

Abbreviations: HL, hyperlipidemia group; Non-HL, non-hyperlipidemia group; k CNY, thousands of Chinese yuan; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; FBG, fasting blood glucose; PBG, postprandial blood glucose; HOMA-IR, homeostasis model assessment of insulin resistance; TG, triglyceride; TC, total cholesterol; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol

Analysis of hyperlipidemic subgroups after PSM

After PSM, 1,914 participants were included in the non-HL group and 1,213 in the HL group, respectively. This balance was visually verified through standardized mean differences in covariates and the distribution of propensity scores.

Similar to Section “Analysis of Hypertensive Subgroups After PSM”, Figs. 4 and 5 illustrate the SMD values of each covariate and the distribution characteristics of propensity scores in the hyperlipidemic subgroups after PSM, respectively. These results confirm that PSM has effectively constructed a study sample with reliably comparable baseline characteristics. Similarly, there were no statistically significant differences in all included baseline characteristics between the two groups.

Fig. 4.

Absolute standard mean differences of each covariate before and after propensity score matching between hyperlipidemic subgroups

Fig. 5.

Comparison of propensity scores before and after propensity score matching between hyperlipidemic subgroups. HL: hyperlipidemic group; Non-HL: non-hyperlipidemic group

However, the HL group still showed significant differences in metabolic characteristics, and the prevalence of hypertension was significantly higher than in the non-HL group (P < 0.001), with a higher rate of antihypertensive drug use (P < 0.001). Metabolically, the HL group maintained significantly higher systolic blood pressure (P < 0.001), diastolic blood pressure (P < 0.001), and BMI (P < 0.001), indicating unresolved metabolic burden related to obesity and blood pressure. The HL group also had more significant insulin resistance, with significantly higher HOMA-IR (P < 0.001), along with significantly elevated triglyceride (P < 0.001) and total cholesterol (P < 0.001) levels, confirming that hyperlipidemia itself continuously affects lipid metabolism and insulin sensitivity.

In contrast, the HL group achieved better glycemic control, with significantly lower postprandial glucose (P < 0.001) and HbA1c (P < 0.001), while HDL-c showed a statistical difference (P = 0.009) (see Table 4).

Table 4.

Demographic and clinical characteristics of hyperlipidemic subgroups after PSM

	Non-HP(n = 1914)	HP(n = 1213)	P
Age(years)	52.7 ± 10.1	53.1 ± 9.6	0.329
Gender(male/female)	1269/645	797/416	0.732
Duration(months)	95.3(48.6,147.9)	95.2(48.6,147.8)	0.907
Income(k CNY/year)	29.5(22.2,36.4)	30.1(22.6,36.8)	0.177
Education > 12years			0.124
Yes (n, %)	985(51.5)	590(48.6)
No (n, %)	929(48.5)	623(51.4)
Smoking status			0.271
Current-smoker (n, %)	347(18.1)	242(20.0)
Occasional-smoker (n,%)	519(27.1)	303(25.0)
Non-smoker (n, %)	1048(54.8)	668 (55.1)
Drinking status			0.099
Current-drinker (n, %)	183(9.6)	142(11.7)
Occasional-drinker (n, %)	815(42.6)	484(39.9)
Non-drinker (n, %)	916(47.9)	587(48.4)
Hypertension			< 0.001
Yes (n, %)	490(25.6)	604(49.8)
No (n, %)	1424(74.4)	609(50.2)
Antihypertensive drug use			< 0.001
Yes (n, %)	377(19.6)	479(39.5)
No (n, %)	1537(80.3)	734(60.5)
SBP(mmHg)	126.6 ± 16.7	130.1 ± 16.5	< 0.001
DBP(mmHg)	77.7 ± 10.4	79.9 ± 10.4	< 0.001
BMI(kg/m2)	24.6 ± 3.5	25.8±0.3.3	< 0.001
FBG(mmol/L)	7.2(6.0,8.7)	7.2(6.1,8.5)	0.845
PBG(mmol/L)	13.9(11.2,17.0)	13.2(10.7,16.1)	< 0.001
HOMA-IR	2.2(1.2,4.1)	2.7(1.5,4.6)	< 0.001
HbA1c(%)	7.8(6.6,9.3)	7.5(6.5,9.0)	< 0.001
TG(mmol/L)	1.5(1.0,2.2)	2.0(1.4,3.0)	< 0.001
TC(mmol/L)	4.3 ± 1.1	4.6 ± 1.2	< 0.001
HDL-c(mmol/L)	1.0(0.8,1.2)	1.0(0.8,1.2)	0.009
LDL-c(mmol/L)	2.3(1.7,3.0)	2.4(1.8,3.1)	0.521
LDL-c(mmol/L)	2.3(1.7,3.0)	2.4(1.8,3.1)	0.167

Note: PSM, propensity score matching; HL, hyperlipidemia group; Non-HL, non-hyperlipidemia group; k CNY, thousands of Chinese yuan; SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; FBG, fasting blood glucose; PBG, postprandial blood glucose; HOMA-IR, homeostasis model assessment of insulin resistance; TG, triglyceride; TC, total cholesterol; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol

Correlation analysis of metabolic indicators

As shown in Fig. 6, HOMA-IR was significantly positively correlated with TG level (P < 0.001) and negatively correlated with HDL-c level (P < 0.001). HbA1c had a weak positive correlation with TC and LDL-c (both P < 0.05) but no significant association with TG (P ≥ 0.05). FBG levels showed no significant correlation with some lipid indicators (HDL-c) or blood pressure indicators (P ≥ 0.05), but exhibited significant positive correlations with TC, TG, and LDL-c (P < 0.05).

Fig. 6.

Correlation analysis of metabolic indicators. DBP, diastolic blood pressure; SBP, systolic blood pressure; TG, triglyceride; TC, total cholesterol; HDL, high-density lipoprotein cholesterol; LDL, low-density lipoprotein cholesterol; FBG, fasting blood glucose; HOMA-IR, homeostasis model assessment of insulin resistance

Discussion

Based on standardized data from the MMC, this study thoroughly analyzed differences in the control of metabolic indicators among T2DM patients with or without hypertension or dyslipidemia. This study clearly indicates that patients without these comorbidities are at a disadvantage in terms of metabolic indicator control. This finding offers a novel perspective for a deeper understanding of the impact of comorbidities on T2DM management. Next, we conduct an in-depth discussion of this discovery in combination with the existing research evidence.

Metabolic control differences in hypertensive subgroups: intervention advantages driven by comorbidities

In this study, T2DM patients without hypertension (non-HP group), both before and after PSM, exhibited poor glycemic control with unsatisfactory HbA1c and post-prandial blood glucose levels. In terms of blood lipid levels, the control of total cholesterol and LDL-C levels was suboptimal. In contrast, despite having higher blood pressure, a larger body mass index (BMI), and more severe insulin resistance (manifested as an increase in HOMA-IR), patients with hypertension (HP group) had more favorable overall metabolic indicators.

This observation may be associated with differences in the intensity of the clinical interventions provided to patients with comorbidities. The utilization rate of lipid-lowering drugs was significantly higher in the HP group than in the non-HP group (27.6% vs. 9.6% after PSM). These data strongly suggest that when T2DM patients also have hypertension, clinicians tend to initiate more proactive and comprehensive management strategies, including multi-target interventions. For example, in actual clinical practice, due to the co-coexistence of hypertension and T2DM in patients, doctors may pay more attention to the overall assessment of patients’ cardiovascular risks and thus adjust the use of lipid-lowering drugs earlier and more actively to reduce blood lipid levels and the risk of cardiovascular diseases.

Numerous studies have confirmed that comorbid states greatly increase the clinical attention given to patients. The prospective cohort study by Chang et al. showed that patients with T2DM and hypertension had the highest risk of new-onset and recurrent stroke within 1 year of follow-up, and factors such as smoking, abnormal LDL-C/TC levels, and insufficient physical activity were independent risk factors [27]. This is because the condition of comorbid patients is more complex and the potential risks are higher [27, 28], which may prompt clinicians to monitor the disease changes more closely and adjust the treatment plan in a timely manner, a pattern that aligns with the better metabolic control observed in this group.

Additionally, in our study the triglyceride (TG) level remained significantly higher in the HP group than in the Non-HP group after PSM (1.7 vs. 1.6 mmol/L, P = 0.003). This pattern suggests a coupling between hypertension and insulin resistance (IR)/lipid dysregulation, whereby IR may further drive TG elevation and worsen lipid metabolism. Evidence from reviews likewise indicates that IR is one of the shared pathological bases linking hypertension and metabolic abnormalities. Regarding indicator selection, the triglyceride–glucose (TyG) index—an alternative marker of IR—shows a significant positive association with uncontrolled hypertension and arterial stiffness, and outperforms the traditional IR metric HOMA-IR in identifying uncontrolled hypertension among adults with hypertension [29, 30]. This supplements the mechanism of the synergistic effect between insulin resistance, hypertension, and lipid metabolism disorders in this study, emphasizes the necessity of selecting appropriate IR indicators and multi-target interventions in comorbidity management, and supports the conclusion that type 2 diabetic patients with hypertension require comprehensive management of metabolic indicators.

Metabolic characteristics of hyperlipidemic subgroups: intensive management offsets pathological disadvantages

Similar to the trend in the hypertensive subgroup, in the hyperlipidemic subgroup study, T2DM patients without hyperlipidemia (non-HL group) were clearly inferior to patients with hyperlipidemia (HL group) in terms of glycemic control, with poor control of HbA1c and post-prandial blood glucose levels. Although the HL group had higher triglyceride and total cholesterol levels and more significant insulin resistance (HOMA-IR 2.7 vs. 2.2), their glycemic control was better.

An in-depth analysis of the reasons behind this result revealed that the utilization rate of antihypertensive drugs in the HL group was significantly higher than that in the non-HL group (39.5% vs. 19.6% after PSM). These data suggest that hyperlipidemia may be associated with increased clinical attention to cardiovascular risks, which in turn may reflect optimized glycemic management in this subgroup. When patients have hyperlipidemia, doctors are aware of the increased risk of cardiovascular diseases, and good glycemic control is of great significance for reducing cardiovascular risks. Therefore, they will actively adjust their hypoglycemic regimens and strengthen their glycemic management.

From a mechanistic perspective, the metabolic characteristics of the HL group strongly support the “synergistic management of metabolic syndrome” hypothesis. Many studies have shown that hyperlipidemia and hypertension often share a common pathological basis, namely, insulin resistance [31, 32]. Intensive lipid-lowering and antihypertensive treatments may be associated with improved glycemic control, potentially through multiple biological pathways. For example, by improving vascular endothelial function, the reactivity of blood vessels to insulin is enhanced, which helps insulin play a better role in lowering blood sugar [33–35]. Simultaneously, reducing the inflammatory response and interference of inflammatory factors with the insulin signaling pathway can also improve insulin sensitivity, thereby improving glycemic control [36].

In addition, the decrease in HDL-c levels in the HL group (P = 0.009) further confirmed the classic metabolic feature of an inverse association between triglycerides and HDL-c [37]. Generally, when triglyceride levels increase, HDL-C levels tend to decrease. This feature is relatively common in patients with metabolic syndrome, further indicating the complexity of metabolic disorders in HL.

Correlation of metabolic indicators: the central role of insulin resistance

Correlation analysis revealed that HOMA-IR was positively correlated with TG and negatively correlated with HDL-C (both P < 0.001). These results confirm that insulin resistance acts as a core hub linking glucose and lipid metabolism disorders, which is fully consistent with the “obesity-insulin resistance-multiple metabolic abnormalities” axis [30, 38]. Insulin resistance reduces the body’s sensitivity to insulin, impairing its ability to promote glucose uptake and utilization, thereby elevating blood glucose levels. Concurrently, it disrupts the normal regulation of fat metabolism, increasing fat breakdown and triglyceride synthesis while inhibiting HDL-c synthesis and metabolism, leading to decreased HDL-c levels.

Notably, in this study, FBG showed no significant association with most metabolic indicators, whereas HbA1c showed only a weak correlation with cholesterol subfractions. This suggests that long-term glycemic control (reflected by HbA1c levels) may be more strongly associated with cholesterol metabolism than short-term glycemic fluctuations (represented by FBG levels). HbA1c reflects the average blood glucose level over the past 2–3 months, providing a more comprehensive picture of long-term glycemic control, and persistent hyperglycemia may exert a more profound impact on cholesterol metabolism.

These findings provide valuable insights for clinical management. Effectively improving insulin resistance in clinical practice is likely to optimize glucose and lipid metabolism simultaneously. For patients with comorbid hypertension or hyperlipidemia, great attention should be paid to the application of “multi-target intervention” strategies. For example, sodium-glucose cotransporter 2 (SGLT2) inhibitors and glucagon-like peptide-1 (GLP-1) receptor agonists improve insulin resistance and provide cardiovascular protection [20, 39]. Multi-target interventions enable comprehensive management of blood glucose, blood lipids, and cardiovascular risks, thereby enhancing treatment efficacy and patients’ quality of life.

Strengths and limitations

First, we used PSM to effectively balance the baseline confounders, which improved between-group comparability and minimized the influence of confounding variables. Second, data were derived from standardized MMC records, reducing biases caused by variations in management models [25] and ensuring greater reliability. Additionally, the simultaneous analysis of both hypertensive and hyperlipidemic subgroups, a multidimensional approach, significantly strengthened the robustness and persuasiveness of the conclusions.

Despite these valuable findings, this study had several limitations. First, single-center data may restrict the generalizability of the results, making them difficult to extrapolate to T2DM populations in different regions or healthcare settings. Future multicenter studies with broader data collection are needed to validate the universality of these conclusions. Second, detailed information on medication use (e.g., specific types of hypoglycemic drugs) was not included, precluding an in-depth analysis of how different intervention regimens affect outcomes and limiting mechanistic exploration. And the better metabolic profiles in the comorbidity groups may partly reflect more intensive clinical follow-up and therapeutic interventions. In addition, data on T2DM-related chronic complications (e.g., diabetic retinopathy, nephropathy, and neuropathy) were not systematically collected in this study, which limits the analysis of their impact on metabolic control. This is a key limitation to be addressed in future research.Third, the cross-sectional design, which enables comparisons between groups at a single time point, cannot establish a causal relationship between comorbidity status and metabolic control. Therefore, prospective studies with long-term follow-up are required to dynamically observe this relationship and clarify causal associations.

Clinical implications

The results of this study clearly indicate that T2DM patients without comorbid hypertension or hyperlipidemia often have poorer metabolic control due to insufficient clinical attention, forming a “potentially underprioritized group” that urgently requires enhanced management. In clinical practice, physicians should avoid formulating treatment plans based solely on single disease indicators; instead, they must fully consider the patients’ comorbid status and implement individualized, precise interventions.

Conclusion

Under the standardized management of MMC, T2DM patients without hypertension or hyperlipidemia show poorer control of key metabolic indicators (such as HbA1c, postprandial glucose, and lipids) than those with these comorbidities. This may be related to the more active interventions received by the patients with comorbidities. These findings highlight the need for strengthened management of T2DM patients without hypertension or hyperlipidemia, who constitute a “potentially underprioritized group.”

Author contributions

Conceptualization: Shuangyan Lv. Data curation: Jing Pang. Formal analysis: Ting Bai, Xiaogang Bai, Jie Xu. Funding: Shuangyan Lv. Investigation: Jing Pang, Ting Bai. Methodology: Jing Pang, Shuangyan Lv. Project administration: Shuangyan Lv. Resources: Shuangyan Lv. Software: Xiaogang Bai, Jie Xu, Xia Li, Wenjing Ren, Yali Zhu, Chuchu Wan, Yan Liu, Jingjing Liu. Supervision: Shuangyan Lv. Validation: Jing Pang, Ting Bai, Xiaogang Bai, Jie Xu, Xia Li, Wenjing Ren, Yali Zhu, Chuchu Wan, Yan Liu, Jingjing Liu. Visualization: Jing Pang. Writing-original draft: Jing Pang, Ting Bai. Writing-review & editing: Shuangyan Lv.

Funding

This work was supported by the Yan’an Science and Technology Plan Project (No. 2024-SFGG-109).

Data availability

The datasets for this study are not publicly available because of concerns regarding participant/patient anonymity. Requests to access the datasets were directed by the corresponding author.

Declarations

Ethical approval

The authors are accountable for all aspects of this work, ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the principles of the Declaration of Helsinki. Owing to the retrospective nature of the study, this research was approved by the Ethics Committee of Yan’an University Affiliated Hospital (Ethics Approval No.: IIT-R-20250169) and the requirement for informed consent was waived. All patient data were anonymized to maintain confidentiality throughout the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.