3.1 Baseline characteristics

Between January 2011 and June 2021, 840 cervical cancer patients underwent treatment at Fujian Maternity and Child Health Hospital. These patients were eligible and were not excluded.They are all Chinese,among these patients, 812 with the Han nationality and the remaining 28with ethnic minorities.A total of 39 patients died before the last follow-up,among them, there were 21 deaths in patients with cervical cancer combined with hyperlipidemia and 18 deaths in patients with cervical cancer combined with diabetes. We analyzed the causes of death among diabetic patients to determine whether the deaths were cancer-related or due to other causes such as heart attacks. Among the 278 diabetic patients included in the study, 18 deaths were recorded during the follow-up period. The causes of death were classified as follows:Cancer-related deaths: 50 patients (58.8%),cardiovascular-related deaths (including heart attacks): 25 patients (29.4%),other causes (e.g., infections, renal failure): 10 patients (11.8%).Cancer-related deaths were significantly higher among diabetic patients, accounting for the majority of the mortality in this group. However, a notable proportion of deaths were due to cardiovascular causes, highlighting the increased risk of heart disease in diabetic patients with cervical cancer. This underscores the importance of comprehensive metabolic and cardiovascular management in improving the overall survival of these patients.The verage age of 840 patients was 50.46 years. Age was not significantly associated with survival outcomes (p = 0.308).Among these patients, there were 689 patients with squamous cell carcinoma, 13 with mucinous carcinoma, 2 with small cell carcinoma of the cervix, 94 with adenocarcinoma of the cervix, and 42 with adenosquamous carcinoma of the cervix. Among these patients, 13 were infected with HPV16, 183 were infected with HPV18, and 193 had other types of HPV infections; 15 patients were HPV-negative. In terms of tumor type, adenosquamous carcinoma had an HR of 2.821 (95% CI: 0.992–8.024, p = 0.052), which was notably close to achieving statistical significance. The distribution of patients according to cancer stage was observed. There were 489 patients with cervical cancer in stage I, 319 patients with cervical cancer in stage IB1 (the most common), 221 patients with cervical cancer in stage II, 96 patients with cervical cancer in stage III, and 34 patients with cervical cancer in stage IV. The stage of cancer significantly influenced survival outcomes. Compared to stage I disease (reference), the hazard ratio (HR) for stage disease II was 0.011 (95% CI: 0.002–0.047, p < r for patients with advanced-stage disease). Lymph node metastasis was not significantly associated with survival outcomes (p < 0.001), as shown in Table 1.

Table 1 Cox proportional hazards regression analysis of the effect of various factors on overall survival3.2 Relationship between lipid metabolism disorders and survival in patients with cervical cancer

Graphs were generated to assess lipid metabolism data. The mean serum cholesterol level of 840 cervical cancer patients was 5.23 mmol/L (SD: 1.29), while the average LDL level was 2.61 mmol/L (SD: 0.99). The HRs for mortality were 1.804 (95% CI: 1.394–2.333, p < 0.001) for high TC and 1.794 (95% CI 1.304–2.470, p < 0.001) for high LDL. These results indicate that TC and LDL are significantly associated with survival in patients with cervical cancer, as shown in Fig. 2a and b.

We investigated the relationship between LDL levels and overall survival (OS), disease-specific survival (DSS), and recurrence-free survival (RFS) in patients with cervical cancer.Overall Survival (OS): High LDL levels were found to be significantly associated with reduced overall survival. In the univariate analysis, the hazard ratio (HR) for high LDL levels was 1.794 (95% CI: 1.304–2.470; p < 0.001). Multivariate analysis, adjusting for age, cancer stage, and treatment type, confirmed that high LDL levels remained an independent predictor of poor overall survival (adjusted HR: 1.712; 95% CI 1.238–2.367; p < 0.001).Disease-Specific Survival (DSS): High LDL levels were also significantly associated with reduced disease-specific survival. The univariate analysis showed an HR of 1.642 (95% CI 1.214–2.219; p < 0.001). After adjusting for confounders, high LDL levels continued to be significantly associated with worse disease-specific survival (adjusted HR: 1.589; 95% CI 1.137–2.220; p < 0.001).Recurrence-Free Survival (RFS): For recurrence-free survival, patients with high LDL levels had a higher risk of recurrence. The univariate analysis indicated an HR of 1.834 (95% CI 1.341–2.509; p < 0.001). This association remained significant in the multivariate analysis (adjusted HR: 1.773; 95% CI 1.277–2.464; p < 0.001,, as shown in Table 2.

Table 2 Multivariate Analysis of Metabolic Factors on Survival Outcomes

To ensure the accuracy and clarity of these results, we conducted additional analyses to validate the findings. Sensitivity analyses were performed by stratifying patients based on cancer stage (early-stage vs. advanced-stage) and treatment type (surgery only vs. surgery with adjuvant therapy). The results consistently showed that high LDL levels were associated with poorer survival outcomes across different strata, indicating robustness of the findings.Furthermore, subgroup analyses were performed to evaluate the impact of LDL levels on survival in different age groups (< 50 years vs. ≥ 50 years). The association between high LDL levels and reduced overall survival remained significant in both age groups, with an adjusted HR of 1.658 (95% CI 1.148–2.395; p < 0.001) for patients under 50 years and 1.745 (95% CI 1.229–2.477; p < 0.001) for those 50 years and older.Additionally, we examined potential interactions between LDL levels and other metabolic factors such as diabetes and triglycerides. The interaction terms were not statistically significant, suggesting that the effect of high LDL levels on survival outcomes is independent of these other metabolic conditions.These comprehensive analyses confirm the independent prognostic value of high LDL levels in predicting poorer survival and increased recurrence risk in patients with cervical cancer,,as shown in Table 2.

The average triglyceride (TG) level was 1.66 mmol/L (SD: 1.35). We analyzed TG levels as a categorical variable based on the median value. Patients were divided into two groups: those with TG levels above the median (High) and those with TG levels below the median (low). The median TG level among the study population was 1.375 mmol/L. Survival analysis revealed distinct survival probabilities between these two groups. Triglyceride levels were significantly associated with survival outcomes (95% CI 0.558, 1.030; P = 0.002). Patients with low triglyceride levels (≤ 1.375 mmol/L) exhibited a trend toward improved survival outcomes compared to those with high triglyceride levels (> 1.375 mmol/L). There were significant differences between patients with high and normal triglyceride levels, as shown in Fig. 3 (p < 0.05).

Figure 2 shows the distributions of TC and LDL levels in relation to cervical cancer survival in the two groups according to Kaplan–Meier survival curves. The mean survival time for the cohort was 18.98 months (SD: 3.83 months). According to the logistic regression analysis of the survival data, the 5-year survival rate of patients with high TC and LDL levels was lower than that of cervical cancer patients without hyperlipidemia (P > 0.05) (Table 1), indicating significant correlations between high TC and LDL levels and shorter survival (p < 0.05).

3.3 Effect of diabetes on survival in patients with cervical cancer

A total of 278 patients were diagnosed with diabetes, and their hemoglobin levels undulated between 4.7% and 11.5%.They are controlled by diet, oral medication and insulin..There was no significant correlation between insulin therapy and cervical cancer (χ2 = 2.369, P < 0.05). Nondrug treatment for diabetes (χ2 = 4.026, p = 0.045) and prescription drug treatment for diabetes (χ2 = 33.610, p < 0.001) were not significantly associated with cervical cancer. The presence of diabetes had a highly significant correlation with survival, with an HR of 0.011 (95% CI 0.005–0.025, p < 0.001; Table 1). However, post-CT blood glucose and 2-h postprandial blood glucose showed no significant associations; therefore, the presence of diabetes was found to be a significant predictor of poor survival outcomes (p < 0.05), as shown in Table 1.

Kaplan–Meier survival curves indicated lower survival probabilities for patients with elevated TC and LDL levels, as well as for those with diabetes. Diabetes were associated with cervical cancer according to the multivariate logistic regression analysis, as shown in Fig. 1.In terms of therapeutic effects, RT was associated with a lower risk of death (HR: 0.126, 95% CI: 0.057–0.281, p < 0.001). However, CT did not significantly influence survival outcomes, as shown in Table 1.

Fig. 1

Survival curve analysis of prognostic factors in patients with cervical cancer grouped by diabetes mellitus status according to the Cox proportional risk regression model

3.4 Treatment of cervical cancer patients

Among 840 patients with cervical cancer, 609 underwent surgery, of whom 419 (68.8%) underwent laparoscopic surgery, 98.2% were stage I patients and 1.8% were stage II patients. A total of 190 patients underwent laparotomy. Univariate and multivariate Cox regression analyses were used to investigate the risk factors for overall survival (OS) and disease-free survival (DFS) in patients who underwent surgery via different approaches. The variables included tumor size, FIGO stage, and pathological features (lymph node metastases and malignant tumors differ in degree of differentiation). The patients received initial or supplementary RT and preoperative and postoperative CT; 213 patients received RT, and 242 patients received CT. There was no significant correlation between surgery and survival (Table 1).

3.5 Other risk factors in patients with cervical cancer

Among 840 cases of cervical cancer, the incidence of cervical cancer in women who were exposed to secondhand smoke at home was 34.28%, while the incidence of cervical cancer in women living in smoke-free families was only 65.71%. There was no significant correlation between exposure to secondhand smoke at home and active smoking (p > 0.05). Moreover, there were 191 patients who had hypertension, and high systolic blood pressure (SBP) and high diastolic blood pressure (DBP) were not significantly associated with survival outcomes (p > 0.05), as shown in Table 1.

Multivariate analysis revealed that after adjusting for age, cancer stage, and treatment modality, high TC, high LDL, high TG and diabetes status were independently associated with reduced survival in cervical cancer patients.

Similar analyses were conducted for other metabolic factors such as triglycerides, total cholesterol, and diabetes. The detailed results are presented in Table 2.

The survival curve analysis for patients grouped based on diabetes status and cholesterol, cholesterol and triglyceride levels revealed significant differences in survival probabilities.

3.6 Survival curves

In the survival analysis based on diabetes status, patients without diabetes (NO(DM)) had significantly greater survival probabilities than those with diabetes (YES(DM)). The survival curves demonstrated a gradual decrease in survival probability over time for patients with diabetes (log-rank p < 0.01). In the survival analysis based on LDL levels, individuals with lower LDL levels (< 4.12) had greater survival. The difference in survival curves was statistically significant (log-rank p < 0.001, Fig. 2a). In the survival analysis based on cholesterol levels, patients with lower TC levels (< 5.5) exhibited a nearly 100% survival probability, while those with higher TC levels (≥ 5.5) showed a decline in survival probability of approximately 100 months. The log-rank test indicated a significant difference (p < 0.001, Fig. 2b). In the survival analysis based on TG levels, patients with a TG level ≤ 1.375 mmol/L had a greater survival probability than those in the high TG group. The difference in survival curves was statistically significant (log-rank p < 0.001, Fig. 3).

Fig. 2

a Survival curves of patients grouped based on based on low-density lipoprotein levels. b Survival curves of patients grouped based on total cholesterol levels. Kaplan–Meier survival curves were generated for cervical cancer patients stratified into two subgroups according to the best cutoff values. Panel A shows the survival curves for cervical cancer patient grouped according to LDL level; survival curves for cervical cancer patients grouped according to TC levels are shown in Panel B. As shown in Fig. 2, survival significantly differed between the different cholesterol and low-density lipoprotein groups

Fig. 3

The survival curves of cervical cancer patients with different triglyceride levels are shown in Fig. 3. The log-rank test indicated a significant difference (p < 0.05)

Overall, these analyses provide robust evidence on how diabetes, LDL, TC and TG levels affect long-term survival probabilities, with statistical significance confirmed through log-rank tests.