Translate this page into:
Novel Biomarkers for Assessing the Risk of Cardiovascular Disease in Individuals with Thyroid Dysfunction: Non-High-Density Lipoprotein Cholesterol Versus Non-High-Density Lipoprotein Cholesterol/High-Density Lipoprotein Cholesterol Ratio
*Corresponding author: Emmanuel Akokhamen Omon, Department of Medical Laboratory Science, College of Medicine and Health Sciences, Afe Babalola University, Ado Ekiti, Nigeria. omonea@abuad.edu.ng
-
Received: ,
Accepted: ,
How to cite this article: Omon EA, Ajayi OD, Edet OO, Airhomwanbor KO. Novel Biomarkers for Assessing the Risk of Cardiovascular Disease in Individuals with Thyroid Dysfunction: Non-High-Density Lipoprotein Cholesterol Versus Non-High-Density Lipoprotein Cholesterol/High-Density Lipoprotein Cholesterol Ratio. J Card Crit Care TSS. 2026;10:150-8. doi: 10.25259/JCCC_55_2025
Abstract
Objectives:
Both the non-high-density lipoprotein cholesterol to high-density lipoprotein cholesterol-C ratio (NHHR) and non-high-density lipoprotein cholesterol (HDL-C) are important indicators for assessing cardiovascular disease (CVD) risk. Non-HDL-C provides a comprehensive evaluation of all cholesterol particles that could contribute to plaque formation, while NHHR indicates the balance between atherogenic and protective cholesterol. This study aimed to assess and compare NHHR and non-HDL-C as indicators for assessing the risk of CVD in individuals with thyroid disorders.
Material and Methods:
In this case-controlled study, 122 subjects were recruited. The group included 62 individuals diagnosed with thyroid dysfunction and 60 healthy controls matched for gender and age, all aged between 20 and 65 years. Anthropometric variables such as body mass index (BMI), waist circumference (WC), systolic blood pressure (SBP), and diastolic blood pressure (DBP) were measured. In addition, free triiodothyronine (FT3), free thyroxine (FT4), thyroid-stimulating hormone (TSH), total cholesterol (TC), triglycerides (TG), and HDL-C were estimated using standard procedures. Low-density lipoprotein cholesterol (LDL-C), very LDL-C (VLDL-C), nonHDL-C, and other atherogenic indices were calculated. Significant difference was pegged at p < 0.05.
Results:
Mean values of DBP, SBP, WC, BMI, TC, fasting blood glucose, TG, non-HDL-C, TSH, LDL-C, high-sensitive C-reactive protein, VLDL-C, interleukin 6, and other atherogenic indices were significantly higher (p < 0.05) in subjects compared to controls; mean values of HDL-C, FT3, and FT4 were significantly decreased (p < 0.05). Non-HDL-C and NHHR showed a significant positive correlation with TSH and lipid profile parameters (p < 0.05), but a non-significant negative correlation with BMI, FT3, and FT4. The receiver operating characteristic analysis indicated that all atherogenic indices were better at assessing CVD risk than lipid profile parameters, with an area under the curve >0.7. Logistic regression showed that NHHR (odds ratio: 5.46; 95% confidence interval: 4.168–16.798; p < 0.001) was the most significant independent predictor of CVD risk.
Conclusion:
All atherogenic indices examined showed significant correlation with lipid profiles and thyroid hormones in individuals with thyroid dysfunction. However, the NHHR outperformed traditional lipid parameters and other atherogenic indices, making it a more sensitive indicator of CVD risk assessment and dyslipidemia.
Keywords
Atherogenic indices
Cardiovascular disease risk
Dyslipidemia
Non-high-density lipoprotein cholesterol to high-density lipoprotein cholesterol-C ratio
Non-high-density lipoprotein cholesterol
INTRODUCTION
Thyroid disease compounds cardio vascular disease continues to impose a substantial global burden despite decades of progress in prevention and management. Traditional risk assessment models—largely built on age, blood pressure, cholesterol levels, diabetes status measuring thyroid profile have undoubtedly improved outcomes, yet they remain inherently limited.[1,2] These models often fail to capture the biological complexity and individual variability that underlie cardiovascular events, particularly in patients who fall outside classical high-risk categories. As a result, there is a growing need to move beyond conventional frameworks toward more refined and personalized approaches.
In parallel, the discovery of novel biomarkers has expanded our understanding of cardiovascular risk. Indicators of systemic inflammation, endothelial dysfunction, oxidative stress, and metabolic imbalance have provided deeper insight into the mechanisms driving atherosclerosis and vascular injury. However, the challenge lies not in the lack of data, but in its interpretation. The sheer volume and multidimensional nature of these biomarkers make it difficult to integrate them into routine clinical decision-making using traditional statistical methods.
This is where artificial intelligence (AI) is beginning to redefine the landscape. Unlike conventional analytical tools, AI is capable of processing vast and complex datasets, identifying subtle patterns, and uncovering non-linear relationships that would otherwise remain hidden. By integrating biochemical markers, clinical variables, imaging data, and even genomic information, AI-driven models can generate a more comprehensive and individualized assessment of cardiovascular risk.
Thyroid dysfunction is a pathological condition in which the thyroid gland fails to synthesize the appropriate quantity of thyroid hormones.[3] Hypothyroidism is characterized by insufficient secretion and biological action of thyroid hormones.[4] This condition, which can be mild to severe, is thought to afflict between 2% and 15% of the population. On the other hand, hyperthyroidism is defined as excessive production of thyroid hormone by the thyroid gland or other tissues outside the thyroid.[5,6]
Cardiovascular disease (CVD) is a major etiology of morbidity and mortality globally. The risk of CVD is increased by thyroid dysfunction due to its impact on lipid metabolism.[6] Thyroid hormones primarily regulate lipid metabolism. Therefore, thyroid dysfunction can lead to lipid abnormalities, which elevate the risk of CVD.[7,8] Although the connection between dyslipidemia and subclinical hypothyroidism has been the subject of numerous studies,[9,10] markers other than the conventional lipid profile, such as atherogenic indices, have not been evaluated for predicting CVD in patients with thyroid dysfunction. The conventional lipid profile has a limited ability to predict CVD risk in those with thyroid dysfunction.[11] Therefore, a more comprehensive indicator of atherogenic lipoproteins, such as non-high-density lipoprotein cholesterol (HDL-C), and the non-HDL-C to HDL-C ratio (NHHR) may have higher predictive power. Non-HDL-C and NHHR provide a more comprehensive measure of risk by including all atherogenic lipoproteins (non-HDL-C) and their relationship with the cardioprotective component (HDL-C). Nevertheless, it is uncertain if the NHHR or non-HDL-C alone is a more effective biomarker for evaluating CVD risk in individuals with thyroid dysfunction. This knowledge gap makes it more challenging to develop effective risk management and stratification plans for this vulnerable group. This study aimed to assess and compare non-HDL-C and NHHR as indicators and diagnostic markers of CVD risk in patients with deranged thyroid function.
MATERIAL AND METHODS
Study design
In this case-controlled study design, one hundred and twenty-two participants were recruited. This group consisted of 62 individuals diagnosed with hypothyroidism (43 with subclinical and 19 with overt cases), and 60 healthy controls matched individually for gender and age. The subjects within the age range of 20–65 years were recruited from the outpatient department and Endocrinology Unit of Afe Babalola University Multi-System Hospital. Medical, family history, and demographic variables were obtained from each subject using a questionnaire. The Ethics and Health Research Committee at Afe Babalola University Multisystem Hospital in Ado-Ekiti, Ekiti State, Nigeria, granted ethical approval for this study. After providing a detailed description of the study’s objectives, methods, and significance, each participant proceeded to complete the informed consent form. Privacy of patients and confidentiality were maintained, and the information collected was used exclusively for research purposes. The study was carried out in line with the recommended guidelines outlined in the Declaration of Helsinki.
Diagnostic criteria for thyroid dysfunction
The clinical practice guidelines for hypothyroidism in adults, which set 4.5 mIU/L as the maximum limit for thyroid-stimulating hormone (TSH), are supported by the American Association of Clinical Endocrinologists and the American Thyroid Association.[12] For thyroid hormones, the normal range is as follows: Free triiodothyronine (FT3) (1.4– 4.2 pg/mL), TSH (0.28–6.82 µIU/mL), and free thyroxine (FT4) (0.8–2.0 ng/mL) based on the analytical range provided by the assay manufacturer. TSH ≥4.5 mIU/mL is the clinical threshold for hypothyroidism.[13] Reduced fT4 and/or fT3 levels with elevated TSH are the clinical hallmark of primary (overt) hypothyroidism, whereas normal levels of fT4 and/or fT3 with elevated TSH are the hallmark of subclinical hypothyroidism. Normal fT4 and/or fT3 levels with low levels of TSH indicate subclinical hyperthyroidism, whereas high fT4 and/or fT3 with low levels of TSH indicate primary (overt) hyperthyroidism. Identification of euthyroid (healthy control) status occurs when TSH, fT3, and fT4 are all within the defined normal reference range.[7]
Inclusion and exclusion criteria
Participants in this study were those who had been diagnosed with thyroid dysfunction (hypothyroidism), according to variations in their thyroid hormone levels in the blood (subclinical or overt). Conversely, the control group consisted of healthy individuals with no history of hypothyroidism or hyperthyroidism. Individuals with infectious disorders such as hepatitis, tuberculosis, and gastrointestinal infections, as well as those with chronic conditions like liver, kidney, or pancreatic diseases, were excluded from the study. In addition, nursing or pregnant women, individuals with underlying medical conditions such as diabetes, human immunodeficiency virus/acquired immune deficiency syndrome, CVDs, immunosuppressive medications, alcohol consumption, tobacco use, and those who did not provide written consent were equally excluded.
Sample collection and analysis
Anthropometric variables, including body mass index (BMI) and waist circumference (WC), were taken to assess the physical characteristics of the subjects. The measurements were taken twice, and the average reading was recorded to reduce random error. WC was measured by wrapping a flexible tape around the waist over light clothing and recording it in centimeters (cm). The blood samples were collected in simple, additive-free tubes, allowed to coagulate for 15 min at room temperature, and then centrifuged for 10 min at 1000 rpm to separate the serum. The serum samples were then stored at −20°C until laboratory analysis was required.
Commercial kits were used for laboratory testing. An Erba Chem-7 Biochemistry Semi-auto analyzer was used to analyze the lipid profile and fasting blood glucose (FBG) levels. The enzyme-linked immunosorbent assay (ELISA) method was used to measure thyroid function tests (FT3, FT4, and TSH).[7] (AccuBind®, Product Code: 7025-300, Monobind Inc., Lake Forest, CA 92630, USA). Furthermore, interleukin 6 (IL-6) and high-sensitive C-reactive protein (hs-CRP) were assayed using the ELISA (Elabscience Inc.) technique. FBG was analyzed enzymatically employing the glucose oxidase-peroxidase method (Randox Glucose GOD-PAP kit, Cat No. GL3815, Randox Ltd., Crumlin, UK). The blood lipid profile includes several parameters: Total cholesterol (TC), TG, HDL-C, low-density lipoprotein cholesterol (LDL-C), and very LDL-C (VLDL-C). The Cholesterol Oxidase-Peroxidase method (Randox TC kit, Cat No. CH3810, Randox Ltd., Crumlin, UK) was used to measure TC (mg/dL). Polyvinyl sulfonic acid and polyethylene-glycol methyl ether were used in a modified precipitation procedure to quantify HDL-C (mg/dL) (Randox HDL kit, Cat No. CH3811, Randox Ltd., Crumlin, UK). The glycerol phosphate oxidase (GPO) method (Randox GPOPAP kit, Cat No. TR3823, Randox Ltd., Crumlin, UK) was also used to measure triglycerides (mg/dL) enzymatically.[14]The TG/5 formula was used to determine VLDL-C (mg/dL). The formula TC-HDL-C was used to calculate non-HDL-C (mg/dL), and Friedewald’s calculation, LDL-C (mg/dL) = TC−(HDL-C−TG/5), was utilized to compute LDL-C. The following formula was used to determine the NHHR: = Non-HDL-C/HDL-C.[10] The following formula was used to calculate other atherogenic indices: Atherogenic index of plasma (AIP), lipoprotein combined index (LCI), Castelli risk index-I (CRI-I), atherogenic coefficient (AC), remnant cholesterol (RC) and Castelli risk index-II (CRI-II): AIP = log (TG/HDL-C); LCI = (TC × TG × LDL-C)/HDL-C; CRI-I = TC/HDL-C; AC = (TC–HDL-C)/HDL-C; RC (mg/dL) = non-HDL-C-LDL-C, and CRI-II = LDL-C/HDL-C.[11]CVD risk was classified using abnormal lipid ratio values, that is, AC >3.0, CRI-1 >3.5, CRI-2 >3.3, and AIP >0.4,[12] respectively.
Statistical analysis
The Statistical Package for Social Sciences Inc., Chicago, IL, USA, version 25, was used for the statistical analysis. To compare study parameters between subjects and controls, the Student’s t-test and analysis of variance were employed. The data were presented in tables and figures as mean ± standard deviation. The study employed Pearson’s correlation analysis to ascertain the relationship between the individual variables. Univariate logistic regression analysis was carried out to identify the independent variables associated with CVD risk. Receiver operating characteristic (ROC) analysis was used to calculate the area under the curve (AUC), as well as corresponding sensitivity and specificity of the studied lipid markers and atherogenic indices for assessing CVD risk in individuals with thyroid dysfunction. Before parametric analyses were conducted, the normality of the data distribution was tested using the Shapiro-Wilk test. A p < 0.05 was considered statistically significant.
RESULTS
Table 1 contains the socio-demographic variables of the subjects and the control group. Thyroid hormone dysfunction was higher in females (72.6%) when compared to their male counterparts (29.1%). In addition, 53.3% had a family history of thyroid hormone dysfunction, 69.4% had subclinical hypothyroidism, while 30.6% had overt hypothyroidism.
| Variables | Subjects (n=62) (%) | Control (n=60) (%) |
|---|---|---|
| Gender | ||
| Male | 17 (27.4) | 26 (43.3) |
| Female | 45 (72.6) | 34 (56.7) |
| Marital status | ||
| Single | 23 (37.1) | 33 (55.0) |
| Married | 39 (62.9) | 27 (45.0) |
| Educational status | ||
| Primary | 11 (17.7) | 5 (8.4) |
| Secondary | 24 (38.8) | 14 (23.3) |
| Tertiary | 19 (30.6) | 35 (58.3) |
| No formal | 8 (12.9) | 6 (10.0) |
| Occupation | ||
| Self-employed | 16 (25.8) | 22 (36.6) |
| Employed | 28 (45.2) | 19 (31.7) |
| Unemployed | 7 (11.3) | 13 (21.7) |
| Artisans | 11 (17.7 | 6 (10.0) |
| Religion | ||
| Christianity | 49 (79.0) | 51 (85.0) |
| Islam | 13 (21.0) | 9 (15.0) |
| Family history of thyroid hormone dysfunction | ||
| Yes | 33 (53.3) | 0 (0) |
| No | 29 (46.7) | 60 (100) |
| Classification of thyroid dysfunction | ||
| Overt hypothyroidism | 19 (30.6) | - |
| Subclinical hypothyroidism | 43 (69.4) | - |
The results obtained showed that systolic blood pressure, diastolic blood pressure, WC, BMI, TSH, hs-CRP, and IL-6 were statistically significantly higher, while FT3 and FT4 were significantly lower in individuals with subclinical hypothyroidism (p < 0.05) and overt hypothyroidism compared to controls [Figures 1-3]. No statistically significant variation was observed (p > 0.05) in the FT3/FT4 and FT4/FT3 ratio. The TC, TG, LDL-C, VLDL-C, non-HDL, NHHR, CRI-I, CRI-II, AC, AIP, TG/HDL, RC, LCI, and VLDL-C/HDL-C ratio were significantly higher (p < 0.05), while HDL-C was significantly lower in subjects with subclinical and overt hypothyroidism compared to controls (p < 0.05) [Table 2]. NHHR had significant positive correlation with age (r = 0.336, p = 0.001), TSH (r = 0.275, p = 0.024), FT3 (r = 0.249, p = 0.041), TC (r = 0.513, p = 0.000), TG (r = 0.468, p = 0.019), HDL-C (r = 0.462, p = 0.000), LDL-C (r = 0.387, p = 0.000), and VLDL-C (r = 0.374, p = 0.020). However, it had a non-significant negative correlation with FT4 and BMI, respectively [Table 3].
- Anthropometric variables of the subjects and controls. *Values are significantly different from control at p>0.05. BMI: Body mass index, WC: Waist circumference, SBP: Systolic blood pressure, DBP: Diastolic blood pressure.
- Thyroid hormones of the subjects and controls. *Values are significantly different from control at p>0.05. FT4: Free thyroxine, FT3: Free triiodothyronine, TSH: Thyroid-stimulating hormone.
- Mean values of inflammatory cytokines involved in cardiovascular disease. *Values are significantly different from control at p>0.05. hs-CRP: Highly sensitive C-reactive protein, IL-6: Interleukin 6.
| Parameters | Control (n=60) | Subclinical hypothyroidism (n=43) | Overt hypothyroidism (n=19) | p-value* |
|---|---|---|---|---|
| FBG (mg/dL) | 88.86±6.34a | 108.44±4.23b | 106.17±4.07b | 0.033 |
| TC (mg/dL) | 141.32±12.17a | 199.48±10.12b | 196.31±11.86b | 0.006 |
| TG (mg/dL) | 117.27±10.78a | 177.32±8.94b | 174.66±9.16b | <0.001 |
| HDL-C (mg/dL) | 52.38±8.94a | 36.43±4.69b | 37.14±4.38b | 0.013 |
| LDL-C (mg/dL) | 65.49±9.37a | 127.59±8.33b | 124.24±9.24b | <0.001 |
| VLDL-C (mg/dL) | 23.45±4.39a | 35.46±3.08b | 34.93±3.67b | 0.015 |
| Non-HDL-C (mg/dL) | 88.94±9.23a | 163.05±12.73b | 159.17±11.64b | <0.001 |
| RC (mg/dL) | 23.45±3.41a | 35.46±4.63b | 34.93±5.06b | 0.026 |
| NHHR | 1.70±0.41a | 4.48±0.97b | 4.29±0.78b | 0.021 |
| CRI-I | 2.70±0.32a | 5.48±0.62b | 5.28±0.67b | 0.029 |
| CRI-II | 1.36±0.30a | 3.50±0.34b | 3.35±0.51b | 0.037 |
| AIP | 0.35±0.12a | 0.69±0.11b | 0.67±0.13b | 0.028 |
| AC | 1.69±0.30a | 4.48±0.72b | 4.29±0.69b | 0.023 |
| LCI | 20,720.49±4,451.33a | 123,883.81±4,148.91b | 114,697.89±13,865.25b | 0.021 |
| TG/HDL-C | 2.24±0.26a | 4.87±1.06b | 4.70±1.00b | 0.030 |
| VLDL-C/HDL-C | 0.43±0.11a | 0.97±0.23b | 0.94±0.18b | 0.036 |
FBG: Fasting blood glucose, TC: Total Cholesterol, HDL-C: High-density lipoprotein cholesterol, Non-HDL-C: Non-high-density lipoprotein cholesterol, LDL-C: Low-density lipoprotein cholesterol, VLDL-C: Very low-density lipoprotein cholesterol, TG: Triglycerides, NHHR: Non-HDL-C to HDL-C ratio, AIP: Atherogenic index of plasma, AC: Atherogenic coefficient, LCI: Lipoprotein combined index, CRI-I: Castelli Risk Index-I, CRI-II: Castelli Risk Index-II, RC: Remnant cholesterol. *A series of mean values with distinct superscripts indicates a significant difference at p<0.05, i.e., “a” compared with “b” is statistically significantly different at p<0.05 using post hoc Tukey test
| Variables | Non-HDL-C r (p-value) | NHHR r (p-value) |
|---|---|---|
| Age (years) | 0.348 (0.001)** | 0.336 (0.001)* |
| BMI (kg/m2) | −0.158 (0.256) | −0.087 (0.420) |
| WC (cm) | 0.081 (0.329) | 0.091 (0.387) |
| SBP (mmHg) | 0.069 (0.514) | 0.112 (0.299) |
| DBP (mmHg) | 0.128 (0.231) | 0.134 (0.222) |
| FBG (mg/dL) | 0.109 (0.211) | 0.076 (0.414) |
| TSH (µIU/mL) | 0.263 (0.032)* | 0.275 (0.024)* |
| FT3 (pg/mL) | −0.108 (0.217) | 0.249 (0.041)* |
| FT4 (ng/dL) | −0.131 (0.240) | −0.201 (0.178) |
| FT3/FT4 | −0.065 (0.611) | 0.277 (0.026)* |
| TC (mg/dL) | 0.467 (<0.001)** | 0.513 (<0.001)** |
| TG (mg/dL) | 0.278 (0.019) * | 0.468 (<0.001)** |
| HDL-C (mg/dL) | 0.342 (0.002)* | 0.426 (<0.001)** |
| LDL-C (mg/dL) | 0.421 (<0.001)** | 0.387 (<0.001)** |
| VLDL-C (mg/dL) | 0.263 (0.020)* | 0.374 (<0.001)** |
| IL-6 (ng/mL) | 0.201 (0.087) | 0.289 (0.019)* |
| Hs-CRP (mg/L) | 0.198 (0.296) | 0.276 (0.024)* |
BMI: Body mass index, WC: Waist circumference, SBP: Systolic blood pressure, DBP: diastolic blood pressure, TSH: Thyroid-stimulating hormone, FT4: Free thyroxine, FT3: Free triiodothyronine, FBG: Fasting blood glucose, IL-6: Interleukin 6, Hs-CRP: High-sensitive C-reactive protein, TC: Total cholesterol, Non-HDL-C: Non-high-density lipoprotein cholesterol, NHHR: Non-HDL-C to HDL-C ratio, LDL-C: Low-density lipoprotein cholesterol, VLDL-C: Very low-density lipoprotein cholesterol, HDL-C: High-density lipoprotein cholesterol. Data were represented as correlation coefficients (r) and p values. **The correlation is significant at the two-tailed 0.01 level. *The correlation is significant at the two-tailed 0.05 level
The ROC analysis revealed that non-HDL and NHHR outperformed the conventional lipid profile and other atherogenic indices in predicting CVD risk with an AUC of 0.7 [Table 4]. In the logistic regression analysis, NHHR (odds ratio [OR]: 5.46; 95% confidence interval [CI]: 4.168–16.798; p < 0.001) and non-HDL-C (OR: 4.78; 95% CI: 3.774–14.738; p < 0.001) emerged as the most significant independent lipid indices for assessing CVD, with a risk more than five times higher than those with normal levels of these parameters [Table 5].
| Variables | AUC | 95% CI | Sensitivity (%) | Specificity (%) | p-value* |
|---|---|---|---|---|---|
| TC | 0.671 | 0.620–0.762 | 67.5 | 56 | 0.001 |
| TG | 0.658 | 0.601–0.759 | 74.3 | 54 | 0.001 |
| HDL-C | 0.650 | 0.620–0.796 | 66.1 | 53 | 0.011 |
| LDL-C | 0.641 | 0.566–0.715 | 56.7 | 52 | 0.003 |
| VLDL-C | 0.639 | 0.589–0.738 | 53.8 | 51 | 0.001 |
| NON-HDL-C | 0.899 | 0.875–0.936 | 88.7 | 63 | 0.001 |
| NHHR | 0.924 | 0.906–0.985 | 89.6 | 69 | 0.001 |
| RC | 0.721 | 0.717–0.749 | 80.3 | 56 | 0.001 |
| CRI-I | 0.796 | 0.774–0.825 | 85.4 | 59 | 0.001 |
| CRI-II | 0.762 | 0.748–0.796 | 79.2 | 53 | 0.001 |
| AIP | 0.863 | 0.832–0.915 | 84.6 | 58 | 0.001 |
| AC | 0.788 | 0.760–0.820 | 79.3 | 57 | 0.001 |
| LCI | 0.735 | 0.700–0.761 | 86.3 | 58 | 0.001 |
| TG/HDL-C | 0.837 | 0.815–0.886 | 87.2 | 56 | 0.001 |
| VLDL-C/HDL-C | 0.749 | 0.720–0.790 | 79.5 | 52 | 0.001 |
AUC: Area under the curve, CI: Confidence interval, TC: Total Cholesterol, HDL-C: High-density lipoprotein cholesterol, Non-HDL-C: Non-high-density lipoprotein cholesterol, LDL-C: Low-density lipoprotein cholesterol, VLDL-C: Very low-density lipoprotein cholesterol, TG: Triglycerides, NHHR: Non-HDL-C to HDL-C ratio, AIP: Atherogenic index of plasma, AC: Atherogenic coefficient, LCI: Lipoprotein combined index, CRI-I: Castelli Risk Index-I, CRI-II: Castelli Risk Index-II, RC: Remnant cholesterol. *Values of p<0.05 are considered statistically significant
| Variables | β Coefficient | OR | 95% CI for OR | p-value* | |
|---|---|---|---|---|---|
| Lower | Upper | ||||
| TC | 0.074 | 1.06 | 0.961 | 1.182 | 0.326 |
| TG | 0.059 | 1.02 | 0.910 | 1.246 | 0.467 |
| HDL-C | -0.056 | 1.01 | 0.968 | 1.275 | 0.312 |
| LDL-C | 0.328 | 1.34 | 1.100 | 1.568 | 0.039 |
| VLDL-C | 0.209 | 1.12 | 1.052 | 1.357 | 0.238 |
| NON-HDL-C | 2.114 | 4.78 | 3.774 | 14.738 | 0.001 |
| NHHR | 2.336 | 5.46 | 4.168 | 16.798 | 0.001 |
| RC | 1.521 | 2.67 | 2.894 | 11.846 | 0.001 |
| CRI-I | 1.585 | 2.98 | 2.283 | 10.366 | 0.001 |
| CRI-II | 1.546 | 2.72 | 1.921 | 10.850 | 0.001 |
| AIP | 2.018 | 3.15 | 4.789 | 13.130 | <0.001 |
| AC | 1.768 | 3.46 | 3.866 | 14.875 | <0.001 |
| LCI | 1.563 | 2.77 | 2.762 | 11.106 | 0.001 |
| TG/HDL-C | 1.896 | 3.02 | 3.811 | 14.731 | <0.001 |
| VLDL-C/HDL-C | 1.551 | 2.73 | 2.467 | 14.843 | 0.001 |
OR: Odds ratio, CI: Confidence interval, TC: Total cholesterol, HDL-C: High-density lipoprotein cholesterol, Non-HDL-C: Non-high-density lipoprotein cholesterol, LDL-C: Low-density lipoprotein cholesterol, VLDL-C: Very low-density lipoprotein cholesterol, TG: Triglycerides, NHHR: Non-HDL-C to HDL-C ratio, AIP: Atherogenic index of plasma, AC: Atherogenic coefficient, LCI: Lipoprotein combined index, CRI-I: Castelli Risk Index-I, CRI-II: Castelli Risk Index-II, RC: Remnant cholesterol. *Values are statistically significant at p>0.05. Logistic regression analysis was used to examine the data after controlling for gender and age. Dependent variables: NON-HDL-C and NHHR
DISCUSSION
One of the most promising aspects of this approach is its ability to move toward true precision medicine. Rather than assigning patients to broad risk categories, AI enables the creation of personalized risk profiles that evolve over time. Two individuals with similar traditional risk factors may, in fact, have markedly different underlying pathophysiological processes—and therefore different levels of risk. AI has the potential to recognize these distinctions, allowing clinicians to tailor preventive and therapeutic strategies more effectively.
Equally important is the potential for dynamic risk monitoring. Cardiovascular risk is not static; it changes in response to lifestyle, treatment, and emerging comorbidities. AI systems can continuously incorporate new data, offering real-time reassessment and enabling earlier intervention when risk begins to escalate. This shift from episodic evaluation to continuous surveillance represents a fundamental change in how we approach cardiovascular care.
Although thyroid disorders significantly influence lipid levels, knowledge of the impact of atherogenic indices remains limited. This study revealed that serum HDL-C levels were significantly lower in patients with thyroid dysfunction compared to healthy controls. In addition, serum TG, TC, LDL-C, and VLDL-C levels were significantly higher in these patients. This aligns with previous research that has documented significantly higher serum LDL-C, VLDL-C, TG, and TC levels, accompanied by a corresponding decrease in HDL-C, in individuals with thyroid disorders compared to healthy controls.[6,8,14] In addition, individuals with thyroid dysfunction had a significantly higher BMI than controls, indicating a relationship between thyroid function and obesity. This finding aligns with previous studies that have linked obesity, hypothyroidism, and the risk of developing dyslipidemia.[15,16] Furthermore, our findings showed that overt hypothyroidism was connected to the highest levels of TC, TG, and LDL-C, followed by subclinical hypothyroidism with slightly lower values. In agreement with previous studies,[17-19] our findings support the hypothesis that individuals with hypothyroidism have a greater chance of developing CVD due to abnormality in their lipid profile markers.[20] The significantly reduced HDL-C and higher LDL-C levels found in this study suggest that hypothyroidism may be a risk factor in the pathogenesis of atherosclerosis, which demonstrates the diverse impact of thyroid hormones on the lipid profile.[21,22]
In this research, all assessed atherogenic indices were significantly higher in individuals with thyroid dysfunction than in the controls. Univariate logistic regression analysis revealed that elevated levels of various atherogenic indices increase the risk of CVDs. The AUC from the ROC analysis for each atherogenic index exceeded 0.7, indicating good performance. These results corroborate earlier research linking atherogenic indices to CVD risk.[23,24] Elevated LCI is linked to increased TG, TC, and LDL-C levels. Changes in CRI-I and CRI-II suggest dyslipidemia, which may play a role in the development of coronary heart disease.[25] Elevated AC values suggest an increased risk of plaque buildup and subsequent cardiovascular events, along with a greater presence of atherogenic particles.[26,27] Although there is limited data on how AIP affects thyroid function, it is understood that thyroid dysfunction can lead to complex alterations in lipid metabolism, accurately measuring the overall changes in the lipid profile caused by thyroid dysfunction.[28]
When compared to other atherogenic markers, both nonHDL-C and NHHR showed superior performance, despite the fact that each is linked to an increased risk of CVDs. Non-HDL-C achieved an AUC of 0.899, and a sensitivity and specificity of 88.7% and 63%, while NHHR reached an AUC of 0.924, with a sensitivity and specificity of 89.6% and 69%, respectively. The odds ratios for non-HDL-C and NHHR were 4.78 and 5.47, respectively. In agreement with a previous study,[29] our results showed that NHHR performed better when compared to non-HDL-C as a marker of CVD risk in individuals with thyroid dysfunction. We found a significant positive association between non-HDL-C levels and age, TC, TG, LDL-C, and VLDL-C in individuals with thyroid dysfunction, consistent with findings from previous research.[6,30,31] A low-grade inflammatory state, which might impact hormone metabolism, is mostly associated with elevated non-HDL-C levels. Research suggests that inflammatory markers may activate deiodinases, an enzyme that removes iodine from TT4, leading to the production of active TT3 and inactive diiodothyronine.[32] We assessed hs-CRP and IL-6 as markers of inflammation and found that both were significantly elevated in individuals with thyroid dysfunction, further linking thyroid dysfunction to a higher chance of developing CVD.[33]
The NHHR is a valuable tool in clinical settings, essential for assessing the levels and functionality of thyroid hormones.[34] Its clinical significance lies in its strong association with the risk, progression, and prognosis of various metabolic, cardiovascular, and inflammatory diseases, often outperforming traditional individual lipid markers. This study emphasizes the use of NHHR when evaluating CVD in individuals with thyroid dysfunction. NHHR is a recently developed comprehensive lipid marker that integrates both pro-atherogenic (nonHDL-C) and anti-atherogenic (HDL-C) lipid components, demonstrating superior predictive value for CVD events, angina pectoris, and all-cause mortality compared to traditional measures like LDL-C or non-HDL-C alone. After controlling for different confounding variables, NHHR presents as a significant independent risk factor for CVD in individuals with thyroid dysfunction. When compared with other lipid parameters, we discovered that NHHR performed better than conventional lipid profile markers as a novel lipid metabolism index. One possible explanation is that cholesterol is extensively present in mammalian cell membrane structures and is essential for preserving the fluidity, permeability, and microstructure of these membranes.[34] The link between NHHR and thyroid hormones may also be significantly influenced by how thyroid hormones affect the regulation of cholesterol. By preventing the production of hepatic bile acids, TSH can lower cholesterol excretion and aid in maintaining cholesterol balance.[35] Consistent with the results of our study, research has also shown that NHHR is a useful marker for assessing CVD risk.[36] Individuals with hypothyroidism have higher chances of developing CVD, and individuals who undergo heart surgery may also develop hypothyroidism.[37] Using NHHR in clinical practice will improve precision in the assessment and early diagnosis of CVD for patients with thyroid-related diseases.
One of the limitations of this study is the uneven ratio of female to male participants, which could skew the findings. The greater incidence of thyroid problems in women as opposed to men may explain why the study had a higher percentage of female participants. Due to a limited number of cases in the specific geographic area, it was difficult to recruit an adequate number of male participants. Furthermore, the research was conducted in a single location in western Nigeria. It is recommended that a larger sample size, equal representation of both genders, and a multicenter approach be utilized for a more comprehensive investigation.
Future directions
Use of AI with novel biomarkers
Recent advances in AI have opened new avenues for identifying and integrating novel biomarkers to refine cardiovascular risk prediction in individuals with thyroid dysfunction Beyond traditional lipid indices such as nonHDL cholesterol and NHHR, Al-driven models now incorporate multidimensional data-including inflammatory markers (e.g., hs-CRP, IL-6), oxidative stress indicators, metabolomic signatures and genetic profiles to uncover subtle, non-linear associations between thyroid hormone imbalance and cardiovascular pathology. Machine learning algorithms, particularly deep learning and ensemble methods can analyze large-scale datasets to detect early patterns of endothelia dysfunction, subclinical atherosclerosis, and cardiometabolic derangements that may not be apparent through conventional statistical approaches. Emerging studies up to 2026 suggest that Al-assisted biomarker panels significantly enhance risk stratification by enabling personalized prediction models, thereby facilitating earlier intervention and targeted therapeutic strategies in this high-risk population. Importantly, such approaches also allow dynamic risk assessment by continuously integrating longitudinal patient data, marking a shift from static to adaptive cardiovascular risk evaluation in thyroid disorders
CONCLUSION
All atherogenic indices examined showed a significant correlation with lipid profile markers and thyroid hormones in individuals with thyroid dysfunction. Both non-HDL-C and NHHR are important biomarkers for assessing individuals who have thyroid dysfunction. However, the NHHR outperformed traditional lipid parameters and other atherogenic indices, making it a more sensitive indicator of dyslipidemia and CVD risk assessment. The study emphasized the importance of monitoring lipid levels with atherogenic indices to prevent or reduce the incidence of CVD in individuals with thyroid dysfunction.
Ethical approval:
The research/study was approved by the Institutional Review Board at Afe Babalola University Multi-System Hospital (AMSH), number AMSH/REC/25/55/116, dated 28th August 2025.
Declaration of patient consent:
The authors certify that they have obtained all appropriate patient consent forms. In the form, the patient has given consent for clinical information to be reported in the journal. The patient understand that the patient’s names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.
Conflicts of interest:
There are no conflicts of interest.
Use of artificial intelligence (AI)-assisted technology for manuscript preparation:
The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.
Financial support and sponsorship: Nil.
References
- Paradigms of Dynamic Control of Thyroid Hormone Signaling. Endocr Rev. 2019;40:1000-47.
- [CrossRef] [PubMed] [Google Scholar]
- Association between Non-High-Density Lipoprotein Cholesterol to High-Density Lipoprotein Cholesterol Ratio and Serum Thyroid Function Measures: Recent Findings from NHANES 2007-2012 and Mendelian Randomization. Front Endocrinol. 2025;16:1467254.
- [CrossRef] [PubMed] [Google Scholar]
- Oxidative Stress and Antioxidant Markers in Individuals with Thyroid Hormone Dysfunction. Eur J Clin Exp Med. 2023;21:768-75.
- [CrossRef] [Google Scholar]
- Subclinical Hypothyroidism In: StatPearls. Treasure Island, FL: StatPearls Publishing; 2025. Available from: https://www.ncbi.nlm.nih.gov/books/NBK536970 [Last accessed on 2025 Sep 03]
- [Google Scholar]
- Thyroid Hormones and Cardiovascular Function and Diseases. J Am Coll Cardiol. 2018;71:1781-96.
- [CrossRef] [PubMed] [Google Scholar]
- Association of Non-High-Density Lipoprotein Cholesterol With Cardiometabolic Risk Factors in Patients With Hypothyroidism. Cureus. 2024;16:e64531.
- [CrossRef] [Google Scholar]
- Study of the Association between Thyroid Dysfunction and Serum Lipid Abnormalities. Biomed Rep. 2024;21:138.
- [CrossRef] [PubMed] [Google Scholar]
- Assessment of Cardiovascular Disease Risk in Females with Subclinical Hypothyroidism. J Lipids. 2023;11:4440275.
- [CrossRef] [Google Scholar]
- Serum non-HDL A Biomarker for Evaluation of Cardiovascular Risk in Hypothyroid Patients. Int J Med Pub Health. 2025;15:551-4.
- [Google Scholar]
- The Association Between Non-High-Density Lipoprotein Cholesterol to High-Density Lipoprotein Cholesterol Ratio (NHHR) and Suicidal Ideation in Adults: A Population-Based Study in the United States. Lipids Health Dis. 2024;23:17.
- [CrossRef] [Google Scholar]
- Subclinical Hypothyroidism and Symptoms of Depression: Evidence from the National Health and Nutrition Examination Surveys (NHANES) Compr Psychiat. 2021;109:152253.
- [CrossRef] [PubMed] [Google Scholar]
- Pediatric Thyroid Disorders in Two Teaching Hospitals in South-West Nigeria. Niger J Med. 2020;29:428-32.
- [CrossRef] [Google Scholar]
- Associations between Indicators of Lipid and Glucose Metabolism and Hypothyroidism. l. Lipids Health Dis. 2025;24:58.
- [CrossRef] [PubMed] [Google Scholar]
- Thyroid Functions and Serum Lipid Profile in Metabolic Syndrome. Biomed J. 2017;40:147-53.
- [CrossRef] [PubMed] [Google Scholar]
- Relationship of Subclinical Hypothyroidism on Epicardial Adipose Tissue: A Systematic Review and Meta-Analysis. Curr Probl Cardiol. 2023;48:101674.
- [CrossRef] [PubMed] [Google Scholar]
- Focus on Cardiometabolic RISK Factors. Acta Cardiol. 2023;78:515-8.
- [CrossRef] [PubMed] [Google Scholar]
- Lipid Profile Pitfalls in Subclinical Hypothyroidism Pathophysiology and Treatment. Lipidology. 2024;1:105-16.
- [CrossRef] [Google Scholar]
- The Relationship Between Thyroid Hormones and Lipid Profile in Subclinical Hypothyroidism Female Patients. Baghdad J Biochem Appl Biol Sci. 2022;3:200-9.
- [CrossRef] [Google Scholar]
- A Study of Thyroid Profile Status and its Correlation with Lipid Profile in a Tertiary Care Centre. Int J Contemp Med Res. 2024;11:1-4.
- [Google Scholar]
- NAFLD and thyroid function: Pathophysiological and therapeutic considerations. Trends Endocrinol Metab. 2022;33:755-68.
- [CrossRef] [PubMed] [Google Scholar]
- The role of thyroid-stimulating hormone in regulating lipid metabolism: Implications for body-brain communication. Neurobiol Dis. 2024;201:106658.
- [CrossRef] [PubMed] [Google Scholar]
- Correlation of Lipid Profile Parameters with Thyroid Dysfunctions-A Cross-Sectional Analytical Study' CME J Geriatr Med. 2024;16:92-5.
- [Google Scholar]
- Study of Relationship between Abnormal Thyroid Function and Lipid Levels in Subclinical and Overt Hypothyroidism Patients in a Rural Tertiary Care Hospital. Int J Med Sci Innov Res. 2022;7:98-108.
- [Google Scholar]
- Association of Atherogenic Indices and Triglyceride-Total Cholesterol-Body Weight Index (TCBI) with Severity of Stenosis in Patients Undergoing Angiography: A Case-Control Study. BMC Res Notes. 2025;18:180.
- [CrossRef] [PubMed] [Google Scholar]
- Relationship between Plasma Atherogenic Index and Thyroid Function: Evidence from NHANES 2007-2012. Sci Rep. 2025;15:23870.
- [CrossRef] [PubMed] [Google Scholar]
- Atherogenic Index of Plasma: Novel Predictive Biomarker for Cardiovascular Illnesses. Arch Med Res. 2019;50:285-94.
- [CrossRef] [PubMed] [Google Scholar]
- A Comparative Study on Serum Lipid Levels in Patients with Thyroid Dysfunction: A Single-Center Experience in Ethiopia. BMC Endocr Disord. 2025;25:47.
- [CrossRef] [PubMed] [Google Scholar]
- Non-HDL-Cholesterol in Dyslipidemia: Review of the State-of-the-Art Literature and Outlook. Atherosclerosis. 2023;383:117312.
- [CrossRef] [PubMed] [Google Scholar]
- Association Between Non-HDL-C/HDL-C Ratio and Carotid Intima-Media Thickness in Post-Menopausal Women. J Clin Med. 2022;11:78.
- [CrossRef] [PubMed] [Google Scholar]
- Association of Non-HDL-C-toHDL-C Ratio with Coronary Lesions and its Prognostic Performance in first-Onset NSTEMI. Biomark Med. 2023;17:29-39.
- [CrossRef] [PubMed] [Google Scholar]
- Residual Atherosclerotic Cardiovascular Disease Risk: Focus on non-High-Density Lipoprotein Cholesterol. J Cardiovasc Pharmacol Ther. 2023;28:10742.
- [CrossRef] [PubMed] [Google Scholar]
- HDL-C and non-HDL-C Levels are Associated with Anthropometric and Biochemical Parameters. J Vasc Bras. 2019;18:e20180109.
- [CrossRef] [PubMed] [Google Scholar]
- Association Between C-Reactive Protein and Interleukin-6 Levels in Subclinical Hypothyroid Patients. J Datta Meghe Inst Med Scie Univ. 2018;13:195-8.
- [CrossRef] [Google Scholar]
- Exploring the Relationship Between NHHR and the Degree of Coronary Artery Stenosis in Patients with Acute Coronary Syndromes. BMC Cardiovasc Disord. 2025;25:589.
- [CrossRef] [PubMed] [Google Scholar]
- Subclinical Hypothyroidism and Coronary Revascularization after Coronary Artery Bypass Grafting. Am J Cardiol. 2018;122:1862-70.
- [CrossRef] [PubMed] [Google Scholar]
- Triiodothyronine and Free Thyroxine Levels are Differentially Associated with Metabolic Profile and Adiposity-Related Cardiovascular Risk Markers in Euthyroid Middle-Aged Subjects. Thyroid. 2014;24:223-31.
- [CrossRef] [PubMed] [Google Scholar]
- Utility of non-High-density Lipoprotein Cholesterol to High-Density Lipoprotein Cholesterol Ratio in Evaluating Incident Diabetes Risk. Diabetes Metab Syndr Obes. 2022;15:1677-86.
- [CrossRef] [PubMed] [Google Scholar]