Sage Journals: Discover world-class research

Abstract

Objective

Arthritis is an acute or chronic inflammatory joint disease characterized by pain and stiffness that typically worsen with aging. It is often accompanied with joint rigidity, cartilage degeneration, and pathological changes, such as monocyte infiltration, inflammation, synovial swelling, and mass formation. Osteoarthritis and rheumatoid arthritis are the two most common forms of arthritis worldwide. The atherogenic index of plasma, which reflects lipid metabolism, may influence arthritis progression. This study aimed to investigate the relationship between the atherogenic index of plasma and arthritis among adults in the United States.

Methods

Using the National Health and Nutrition Examination Survey data (2005–2018), 12,273 individuals (representing 74.2 million people) with complete data on the atherogenic index of plasma and arthritis assessments were analyzed. The atherogenic index of plasma was calculated using the following formula: log₁₀(triglycerides/high-density lipoprotein cholesterol). Weighted multivariable logistic regression, restricted cubic spline, and subgroup analyses were performed to evaluate the associations.

Results

Higher atherogenic index of plasma quartiles were correlated with increased arthritis prevalence (p < 0.05). After adjustments, the highest quartile showed a 45% higher arthritis risk (odds ratio = 1.45, 95% confidence interval: 1.20–1.74, p < 0.001) than the lowest quartile. Restricted cubic spline analysis confirmed a linear association between the atherogenic index of plasma and arthritis (p < 0.001). Subgroup analyses revealed significant interactions among age, sex, smoking status, educational level, and race (p = 0.019).

Conclusion

The atherogenic index of plasma was positively correlated with arthritis incidence, highlighting it as a potential biomarker for risk assessment and prevention. Clinicians should monitor elevated atherogenic index of plasma in high-risk populations. Prospective studies should be conducted to validate the causal role of the atherogenic index of plasma and explore its utility in arthritis prevention protocols.

Keywords

Atherogenic index of plasma biomarkers arthritis National Health and Nutrition Examination Survey lipid metabolism

Introduction

Arthritis is an acute or chronic inflammatory joint disease characterized primarily by joint pain and stiffness that worsen with aging. Arthritis is often accompanied with joint rigidity and cartilage degeneration, along with various pathological features, such as monocyte infiltration, inflammation, synovial swelling, and mass formation.^1,2 Osteoarthritis (OA) and rheumatoid arthritis (RA) are the two most common forms of arthritis globally.³ According to the latest 2017 Global Burden of Disease Study, the prevalence of OA exceeds 300 million, with an incidence of approximately 15 million, whereas the age-standardized prevalence of RA has increased globally by 7.4%.⁴ The incidence of OA is expected to increase annually with an aging global population and rising obesity rates.⁵ Statistics has shown that knee and hip arthritis rank as the 11th leading cause of disability worldwide, posing a significant economic burden on the society.⁶ A study on disease costs revealed that the annual economic burden for OA accounts for approximately 0.25%–0.50% of a country’s gross domestic product; this translates into an average annual incremental cost of US$7000 per person, including direct and indirect costs.⁷ In addition, the burden of RA has increased exponentially over time. In terms of socioeconomic burden, in addition to the direct medical costs, the consequences of functional disability and the resulting decrease in work capacity and social participation should be considered.⁸ Therefore, early identification and intervention in arthritis are important to alleviate patient suffering.

The atherogenic index of plasma (AIP) is a novel biomarker calculated as log₁₀(triglycerides (TG)/high-density lipoprotein cholesterol (HDL-C)), which is related to the lipoprotein particle size and may serve as an independent cardiovascular risk factor.^9,10 Zhang et al. found that HDL-C level is negatively correlated with cartilage damage, radiological severity, and severity of primary knee OA (KOA) symptoms, suggesting that it is a predictor of OA severity.¹¹ A cohort study in China discovered that for each unit increase in TG levels, the prevalence and incidence of clinical KOA increased by 9% and 5%, respectively.¹² Furthermore, a study by Iranian scholars investigating the relationship between metabolic syndrome and OA incidence found that serum TG concentrations were significantly higher in patients with OA than in healthy individuals.¹³ A study by Rodríguez-Carrio et al. in Spain on high TG and low HDL-C levels in RA revealed that a TG^highHDL-C^low profile was associated with systemic inflammation in RA and was negatively correlated with clinical outcomes of tumor necrosis factor-α blockade treatment, indicating that a TG^highHDL-C^low profile serves as a viable serum biomarker for assessing clinical response in RA.¹⁴

Therefore, we hypothesized a potential correlation between the AIP and arthritis occurrence. This study aimed to investigate this relationship using a cross-sectional survey database created by the National Center for Health Statistics (NCHS) in the United States (US) to identify a new biomarker for predicting arthritis.

Materials and methods

Data source and study participants

Data from the National Health and Nutrition Examination Survey (NHANES, 2005–2018) were selected to ensure a sufficient sample size for subgroup analyses and include the most recent nationally representative data for this cross-sectional observational study. The NHANES is a cross-sectional survey initiated by the NCHS to collect comprehensive health and nutrition data on the US population, including demographic information, socioeconomic status, dietary habits, health-related questions, and medical history evaluated using physical examinations and laboratory tests (https://www.cdc.gov/nchs/nhanes/index.htm). To obtain a representative sample of research participants, the organization employed a sampling method utilizing a multistage stratified cluster probability sampling approach. The NHANES received ethical approval from the NCHS Institutional Review Board, and all participants provided informed consent. This study reviewed participant information across 7 cycles of NHANES from 2005 to 2018 and included a total of 70,190 participants. After excluding individuals aged <20 years, 39,749 participants remained. Further exclusion criteria ruled out 22,238 individuals who did not complete the arthritis questionnaire, 176 who lacked AIP data, and 5062 without covariate data, resulting in a final sample size of 12,273 (Figure 1).

Figure 1.

Flowchart of the sample selection from NHANES 2005–2018. NHANES: National Health and Nutrition Examination Survey.

Definition of the AIP

The AIP was originally developed based on a study that included 1433 patients with varying degrees of atherosclerosis. The analysis showed a strong positive correlation (r = 0.803) between the indirect measurement of lipoprotein particle size and log₁₀(TG/HDL-C), indicating that log₁₀(TG/HDL-C) serves as a marker of plasma atherogenicity. The term AIP was first introduced in this study.¹⁰ In recent years, AIP has been closely associated with musculoskeletal disorders, such as sarcopenia and abnormal bone density.^15,16 AIP was calculated using blood sample indicators, specifically TG and HDL-C, with rigorous quality control measures to ensure data reliability. AIP was defined as the logarithmic ratio of TG to HDL-C in mmol/L (log₁₀(TG/HDL-C)) and was categorized into quartiles for analysis.

Diagnosis of arthritis

Arthritis prevalence was assessed using self-report questionnaires. A study¹⁷ demonstrated an 85% consistency between self-reported and clinically diagnosed arthritis, indicating the reliability of self-reporting. American adults were asked the following question: “Has a doctor or other health professional ever told you that you have arthritis?” Those responding “yes” were further asked to specify the type of arthritis. The types of arthritis were defined as follows: OA or degenerative arthritis, RA, psoriatic arthritis, and other types. Participants identified as having arthritis were classified into the arthritis group, while others were included in the non-arthritis group.¹⁸

Other covariables

To evaluate the association between the AIP and arthritis, the model was adjusted for the following covariables: age, sex, race, body mass index, waist circumference, alcohol consumption frequency, smoking status, educational level, poverty strata, total cholesterol, low-density lipoprotein cholesterol (LDL-C), HDL-C, TG, albumin, creatinine (CRE), glycated hemoglobin, glucose, and arthritis type (OA or degenerative arthritis, RA, psoriatic arthritis, and other). Alcohol consumption frequency was categorized as 1–5 times per month, 5–10 times per month, >10 times per month, or nondrinkers. Based on smoking status, the participants were classified as current smokers, former smokers, or nonsmokers. Furthermore, educational level was classified as follows: <9th grade, 9th–11th grade, high school graduate, and college or above. Arthritis types included OA, RA, and other forms. Detailed information on these variables can be found on the Centers for Disease Control (CDC) website (www.cdc.gov/nchs/nhanes/).

Statistical analyses

Data were processed using R version 4.3.3. Weighted analyses were performed using the R package survey, and restricted cubic spline (RCS) models were fitted using the rms package. NHANES data analysis adhered to the NHANES statistical guidelines, wherein sample weights were applied for complex multistage probability sampling. The weight variable WTMEC2YR was scaled to 1/7 for the analysis, and all analyses used adjusted weights. Continuous variables with a normal distribution were reported as medians and interquartile ranges. Student t-test was used for comparisons of normally distributed variables, while the Mann–Whitney U test was applied for non-normally distributed variables. To examine the relationship between the AIP and arthritis, weighted logistic regression was performed using AIP quartiles. The following three models were applied: 1. Model 1 employed weighted logistic regression with an unadjusted model, where AIP quartiles served as independent variables and arthritis served as the dependent variable. (b) Model 2 was adjusted for age, sex, and alcohol consumption frequency. 3. Model 3 was further adjusted for age, sex, alcohol consumption frequency, smoking status, poverty strata, race, and educational level.

Results

Baseline characteristics of the participants

Participants aged <20 years were excluded based on the epidemiological characteristics of arthritis. The demographic characteristics of the participants are shown in Table 1. After applying complex weighting, the overall prevalence of arthritis was 26.52% (n = 19,681,517/74,225,012) (Table S1). We found that arthritis was more common in older adults aged ≥60 years (55.22%); moreover, women had a higher prevalence of arthritis than men (59.65% vs. 40.35%). Non-Hispanic Whites had a higher prevalence of arthritis than other ethnic groups (80.31% vs. 19.69%). The proportion of individuals with obesity was higher in the arthritis group than in the non-arthritis group (47.18% vs. 32.58%). Notably, the proportions of both current smokers and nonsmokers were lower in the arthritis group than in the non-arthritis group, whereas the proportion of former smokers was significantly higher in the arthritis group (34.34% vs. 44.08%, p < 0.05). Additionally, significant differences were observed between the arthritis and non-arthritis groups in terms of educational level, HDL-C levels, and CRE levels. However, no significant associations were found between arthritis and poverty or LDL-C levels.

Table 1.

Demographics and characteristics of study participants from NHANES 2005–2018.

Variables	Overall	Non-arthritis	Arthritis	p-value
Variables	N = 74,066,873 (100%)	n = 54,385,356 (73%)	n = 19,681,517 (27%)	p-value
Age, n (%)^b				<0.001
20–39 years	24,665,173 (33%)	23,084,390 (42%)	1,580,783 (8.0%)
40–59 years	27,055,594 (37%)	19,823,898 (36%)	7,231,696 (37%)
60+ years	22,346,107 (30%)	11,477,068 (21%)	10,869,038 (55%)
Sex, n (%)^b				<0.001
Female	37,682,080 (51%)	25,942,782 (48%)	11,739,298 (60%)
Male	36,384,793 (49%)	28,442,574 (52%)	7,942,220 (40%)
Race/ethnicity, n (%)^b				<0.001
Other/multiracial	4,852,908 (6.6%)	4,003,877 (7.4%)	849,031 (4.3%)
Other Hispanic	3,658,209 (4.9%)	3,070,610 (5.6%)	587,599 (3.0%)
Mexican American	5,819,509 (7.9%)	5,064,890 (9.3%)	754,619 (3.8%)
Non-Hispanic Black	7,371,598 (10.0%)	5,687,386 (10%)	1,684,212 (8.6%)
Non-Hispanic White	52,364,650 (71%)	36,558,594 (67%)	15,806,056 (80%)
BMI (kg/m²)^b				<0.001
Underweight (<18.5)	1,186,565 (1.6%)	938,074 (1.7%)	248,491 (1.3%)
Normal (18.5 to <25)	21,100,589 (28%)	16,855,397 (31%)	4,245,192 (22%)
Overweight (25 to <30)	24,775,577 (33%)	18,873,271 (35%)	5,902,306 (30%)
Obese (≥30)	27,004,143 (36%)	17,718,614 (33%)	9,285,528 (47%)
Waist circumference (cm)^a	98 (88, 109)	96 (86, 107)	103 (93, 115)	<0.001
Alcohol consumption frequency, n (%)^b				<0.001
1–5 drinks/month	36,782,719 (50%)	27,426,581 (50%)	9,356,138 (48%)
5–10 drinks/month	7,075,047 (9.6%)	5,531,614 (10%)	1,543,433 (7.8%)
>10 drinks/month	13,068,338 (18%)	9,553,445 (18%)	3,514,893 (18%)
Nondrinker	17,120,389 (23%)	11,856,427 (22%)	5,263,962 (27%)
Smoking status, n (%)^b				<0.001
Current smoker	15,190,398 (21%)	10,942,808 (20%)	4,247,591 (22%)
Former smoker	19,123,049 (26%)	12,364,421 (23%)	6,758,628 (34%)
Nonsmoker	39,753,426 (54%)	31,078,128 (57%)	8,675,299 (44%)
Educational level, n (%)^b				<0.001
Less than 9th grade	3,937,626 (5.3%)	2,639,595 (4.9%)	1,298,031 (6.6%)
9th–11th grade	8,104,600 (11%)	5,603,776 (10%)	2,500,824 (13%)
High school graduate/ GED certificate	16,696,412 (23%)	12,030,650 (22%)	4,665,762 (24%)
Some college or AA degree	23,116,703 (31%)	16,741,716 (31%)	6,374,987 (32%)
College graduate or above	22,211,532 (30%)	17,369,619 (32%)	4,841,913 (25%)
PIR, continuous, n (%)^a	3.02 (1.50, 5.00)	3.09 (1.52, 5.00)	2.84 (1.46, 5.00)	0.300
PIR, strata, n (%)^b				0.700
Above poverty (≥1.3)	58,706,495 (79%)	43,178,247 (79%)	15,528,248 (79%)
Below poverty (<1.3)	15,360,378 (21%)	11,207,109 (21%)	4,153,269 (21%)
TC (mg/dL)^a	191.00 (165.00, 219.00)	190.00 (165.00, 217.00)	193.00 (167.00, 223.00)	0.006
LDL-C (mg/dL)^a	112.00 (90.00, 136.00)	112.00 (90.00, 137.00)	112.00 (88.00, 136.00)	0.600
TG (mg/dL)^a	104.00 (72.00, 154.00)	101.00 (69.00, 148.00)	115.00 (81.00, 168.00)	<0.001
HDL-C (mg/dL)^a	52.00 (43.00, 63.00)	52.00 (43.00, 63.00)	52.00 (43.00, 64.00)	0.085
ALB (mg/L)^a	7.30 (4.10, 14.40)	7.10 (4.00, 13.40)	8.20 (4.50, 17.60)	<0.001
CRE (mg/dL)^a	111.00 (67.00, 166.00)	114.00 (68.00, 171.00)	103.00 (64.00, 151.00)	<0.001
HbA1c (%)^a	5.40 (5.20, 5.70)	5.40 (5.10, 5.70)	5.60 (5.30, 6.00)	<0.001
GLU (mmol/L)^a	5.50 (5.11, 6.00)	5.44 (5.11, 5.88)	5.72 (5.27, 6.38)	<0.001
Arthritis type^b				<0.001
Other	2,629,756 (18%)	0 (NA)	2,629,756 (18%)
Rheumatoid arthritis	3,764,552 (26%)	0 (NA)	3,764,552 (26%)
Osteoarthritis	7,894,932 (55%)	0 (NA)	7,894,932 (55%)
AIP^a	0.30 (0.09, 0.53)	0.28 (0.07, 0.51)	0.34 (0.12, 0.56)	<0.001
Quartiles of AIP^b				<0.001
Quartile 1	18,529,959 (25%)	14,372,612 (26%)	4,157,346 (21%)
Quartile 2	18,512,035 (25%)	13,778,770 (25%)	4,733,265 (24%)
Quartile 3	18,514,386 (25%)	13,285,086 (24%)	5,229,300 (27%)
Quartile 4	18,510,494 (25%)	12,948,888 (24%)	5,561,606 (28%)

Student’s t-test (unpaired) for continuous variables; mean ± SD for continuous variables.

χ² test for categorical data; n (%) for categorical variables.

Weighted t-tests for complex survey samples; chi-square test correction.

Bold values indicate p-value <0.05

NHANES: National Health and Nutrition Examination Survey; BMI: body mass index; GED: General Educational Development; AA: Associate of Arts; PIR: poverty–income ratio; TC: total cholesterol; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; ALB: albumin; CRE: creatinine; HbA1c: glycated hemoglobin; GLU: glucose; AIP: atherogenic index of plasma.

Associations between the AIP and arthritis

All models indicated a positive correlation between the AIP and arthritis. In the crude model (odds ratio (OR): 1.61, 95% confidence interval (CI): 1.38–1.87), after adjusting for all covariates, the likelihood of developing arthritis increased by 64% for each unit increase in AIP (OR: 1.64, 95% CI: 1.35–2.01). When considering AIP as a categorical variable, individuals in the highest AIP quartile had a higher likelihood of arthritis than those in the lowest quartile (this was consistent across all three models). Additionally, the observed p for trend indicated that the increasing trend observed in all models was statistically significant, suggesting that the risk of OA increased with a higher AIP. Based on the comparison of the three models, although the p values were all <0.001, Model 3 had an AIC of 11,867, indicating that Model 3 better illustrated the relationship between the AIP and arthritis (Table 2). Additionally, we performed a comparative analysis of the receiver operating characteristic curve and area under the curve for AIP and OA (Figure 2). Subsequently, we conducted an RCS analysis to explore the nonlinear association between the AIP and arthritis. The fitted RCS curve showed that as the AIP values increased, the OR for OA also increased, which is consistent with the results of multiple logistic regression analysis (Figure 3).

Table 2.

Associations between AIP and arthritis.

Characteristic	Model 1		Model 2		Model 3		Model 4
Characteristic	OR (95% CI)	p-value	OR (95% CI)	p-value	OR (95% CI)	p-value	OR (95% CI)	p-value
AIP (per 1 SD increase)	1.61 (1.38, 1.87)	<0.001	1.85 (1.52, 2.24)	<0.001	1.64 (1.35, 2.01)	<0.001	1.64 (1.35, 2.01)	<0.001
Quartiles of AIP
Quartile 1	Reference		Reference		Reference		Reference
Quartile 2	1.19 (1.02, 1.39)	0.031	1.22 (1.02, 1.46)	0.032	1.15 (0.96, 1.37)	0.132	1.22 (0.99, 1.49)	0.060
Quartile 3	1.36 (1.17, 1.59)	<0.001	1.4 (1.18, 1.66)	<0.001	1.31 (1.10, 1.55)	0.003	1.42 (1.13, 1.78)	0.003
Quartile 4	1.48 (1.27, 1.73)	<0.001	1.62 (1.35, 1.94)	<0.001	1.45 (1.20, 1.74)	<0.001	1.62 (1.23, 2.13)	<0.001
p for trend	<0.001		<0.001		<0.001		<0.001
AIC	14,169		12,117		11,867		11,739
No. Obs.	12,273		12,271		12,271		12,183

AIP: atherogenic index of plasma; OR: odds ratio; CI: confidence interval; AIC: Akaike information criterion.

No. Obs. refers to the number of observations included in each model after accounting for missing data in the covariates.

Data are presented as ORs, 95% CIs, and p-value

Model 1: unadjusted

Model 2: adjusted for age, sex, and alcohol consumption frequency

Model 3: adjusted for age, sex, alcohol consumption frequency, smoking status, poverty strata, race, and educational level

Model 4: adjusted for age, sex, alcohol consumption frequency, smoking status, poverty strata, race, educational level, total cholesterol, high-density lipoprotein, creatinine, and glucose.

Figure 2.

ROC curves for predicting AIP and arthritis. Model 1: unadjusted. Model 2: adjusted for age, sex, and alcohol consumption frequency. Model 3: adjusted for age, sex, alcohol consumption frequency, smoking status, poverty strata, race, and educational level. Model 4: adjusted for age, sex, alcohol consumption frequency, smoking status, poverty strata, race, educational level, total cholesterol, high-density lipoprotein, creatinine, and glucose. Model 5: adjusted for age, sex, alcohol consumption frequency, smoking status, poverty strata, race, educational level, total cholesterol, high-density lipoprotein, creatinine, glucose, and BMI. ROC: receiver operating characteristic; AIP: atherogenic index of plasma; BMI: body mass index.

Figure 3.

The RCS curve fitting the relationship between AIP and arthritis shows that the nonlinear correlation between AIP and the incidence of arthritis is not significant (p = 0.206). The red solid line represents the estimated value, and the pink area indicates the corresponding 95% confidence interval. The associations were adjusted for age, sex, alcohol consumption frequency, smoking status, poverty strata, race, and educational level. RCS: restricted cubic spline; AIP: atherogenic index of plasma.

Subgroup analysis

Overall, the interaction analysis indicated that the correlation between the AIP and risk of arthritis exhibited significant differences across multiple subgroups, including those stratified by sex, age, smoking status, race, and educational level. However, no significant differences were observed between alcohol consumption frequency and the poverty–income ratio (PIR) strata subgroups. When considering AIP as quartiles, in the sex subgroup, the arthritis risk in women increased with higher AIP quartiles, with an OR of 1.74 for Q4, which was significantly higher than that for Q2 (1.25), demonstrating a more pronounced effect of AIP on women. Young participants (aged 20–39 years) showed a significant increase in risk at high AIP levels, with ORs of 1.71 and 1.52 in Q2 and Q4, respectively, indicating a strong impact of AIP on this age group. Regarding smoking status, the arthritis risk for nonsmokers significantly increased with the rising AIP levels (Q4 OR = 1.20), whereas the risk difference for current smokers was not significant, suggesting a minimal impact of AIP within smoking populations. In the other/multiracial and Mexican American groups, an increase in AIP was significantly associated with arthritis risk, particularly in Q4, where ORs were 1.18 and 1.13, respectively, reflecting the moderating effect of race on the relationship between the AIP and arthritis risk. Higher educational levels (college or associate degrees) were correlated with a significant increase in arthritis risk at high AIP levels (Q4 OR = 1.83), suggesting that educational background enhances the influence of AIP on arthritis risk (Figure 4).

Figure 4.

Subgroup analysis of the association between AIP and arthritis. AIP: atherogenic index of plasma; Q1/2/3/4: quartile of atherogenic index of plasma 1/2/3/4.

Discussion

Unlike prior studies linking individual lipid parameters to arthritis, our study established AIP as a novel composite biomarker reflecting the synergistic effects of the TG/HDL-C ratio, providing superior risk stratification. The main objective of this extensive cross-sectional analysis of 12,273 participants from the NHANES was to reveal the correlation between the AIP and arthritis. After adjusting for confounding factors, including age, sex, alcohol consumption frequency, smoking status, poverty strata, race, and educational level, we conducted weighted multivariable logistic regression. This study highlighted a positive correlation between the AIP and arthritis incidence in US adults. Participants in the highest AIP quartile were 45% more likely to develop arthritis than those in the lowest quartile, emphasizing the potential clinical application of AIP as a biomarker for arthritis and offering new insights into the evaluation and treatment of arthritis. We performed RCS analysis to examine the nonlinear relationship between the AIP and arthritis. The fitted RCS curve indicated that higher AIP values were correlated with increased ORs for OA, consistent with the findings from the multiple logistic regression analysis, which revealed significant differences in the association between the AIP and arthritis risk across various subgroups, stratified by sex, age, smoking status, race, and educational level. In contrast, no significant differences were observed for different alcohol consumption frequencies and PIRs. The risk of arthritis in women and young individuals aged 20–39 years increased at higher AIP quartiles. Nonsmokers exhibited a significantly increased risk at high AIP levels, whereas smokers exhibited no significant changes in risk. In the other/multiracial and Mexican American groups, elevated AIP levels were significantly associated with an increased risk of arthritis. Additionally, individuals with higher educational levels showed a significant increase in arthritis risk at high AIP levels, suggesting that educational background augments the impact of AIP on arthritis risk. Clinicians should pay attention to patients with elevated AIP levels, as this indicates a high risk of arthritis incidence, providing valuable implications for arthritis prevention.

Previous studies have shown that many clinical lipid markers, such as cholesterol, cholesteryl esters, free cholesterol, and phospholipids, are closely associated with C-reactive protein. Lipid-related metabolites are also associated with arthritis.¹⁹ AIP, an emerging lipid biomarker, has demonstrated unique value in cardiovascular diseases, metabolic disorders, and inflammatory diseases.^9,20–22 AIP combines the levels of TG and HDL-C and reflects the particle size of lipoproteins, making it a more accurate indicator of the specificity and pathogenicity of lipid abnormalities. Specifically, AIP effectively reflects the levels of small, dense, and cholesterol-rich lipoprotein particles in the blood, which are closely associated with the formation of atherosclerosis. Owing to their high permeability, these small and dense particles easily accumulate within vessel walls, triggering local inflammatory responses and subsequently accelerating the development of atherosclerosis. As the core pathological process of cardiovascular disease, atherosclerosis shares many risk factors with arthritis, particularly the core mechanism of chronic inflammation. Furthermore, increasing evidence suggests a biological link between lipids and bone metabolism. As per previous studies, lipid metabolism disorders can directly affect bone formation and resorption, thereby influencing bone homeostasis.^9,23 At the microscopic cellular level, orthopedic diseases, such as OA and osteoporosis, are pathological conditions caused by disruption of the balance between osteoblasts and osteoclasts. As a marker of lipid levels and inflammatory status, AIP may exacerbate the pathological progression of arthritis via its effects on oxidative stress and endothelial function. As endothelial function deteriorates, the inflammatory response in the blood vessels is amplified, potentially worsening the inflammatory state and the extent of joint damage.^24–26 Evidence linking AIP to various inflammatory and metabolic disorders suggests that AIP is not merely a single lipid indicator; it reflects the interaction of multiple factors, including lipid metabolism, inflammatory response, and oxidative stress. By capturing these complex pathological mechanisms, our study revealed the significance of AIP in the pathological progression of arthritis.

Our study has several strengths that support the robustness and relevance of our findings. First, we employed a large-sample design and conducted a comprehensive analysis of a nationally representative cohort of US adults. This design enhances the external validity and generalizability of our results across diverse populations. Second, we utilized RCS analysis, which allowed us to capture and effectively demonstrate the nonlinear relationship between AIP and arthritis, thus offering deeper insights into the nuanced interactions between these variables. Third, by meticulously adjusting for a wide range of potential confounders, our findings were better aligned with real-world conditions, ensuring that our conclusions are more reflective of actual clinical and epidemiological settings. These methodological strengths collectively enhance the interpretive power and applicability of our research. Fourth, in our subgroup analysis, the association between AIP and OA occurrence was significant across a range of demographic and clinical subpopulations. The consistency within these subgroups further supports the strong association between AIP and OA, which was not significantly altered by these variables. However, there were significant interactions between AIP and subgroups stratified by age, sex, race, educational level, and smoking status, suggesting that these factors mediate the relationship between AIP and OA incidence. Smoking can lead to changes in lipid metabolism, and evidence indicates that smoking is associated with endothelial function and early atherosclerotic mechanisms.²⁷ Additionally, the current epidemiological research suggests that unhealthy lifestyle habits, such as smoking, may affect the occurrence of OA.²⁸ In our study, we found a significant difference in the relationship between AIP and OA among nonsmokers, while it was insignificant among current smokers. This discrepancy may stem from the multifaceted impact of smoking on cardiopulmonary function, which may obscure a clear association between them. Individuals who quit smoking are often aware of the harmful effects of smoking and tend to adopt healthier lifestyle habits, resulting in improved lipid metabolism and cardiovascular health, thereby partially offsetting the effect of smoking on the relationship between AIP and OA.²⁹ Regarding education, multicountry cohort studies have found a significant association between educational level and dyslipidemia occurrence,^30,31 which may be related to differences in dietary patterns among residents with varying levels of education. Further research incorporating dietary patterns is warranted to substantiate this hypothesis.

The current study had certain limitations. First, although our study was based on high-quality standardized surveys, it had a cross-sectional design. Our findings could only demonstrate a linear association between AIP and increased incidence of arthritis, whereas the causal relationship between AIP and increased incidence of arthritis requires further prospective studies for confirmation. Second, our analysis was conducted in a US population. Therefore, it remains to be determined whether these conclusions are applicable to populations in other countries and unincorporated ethnic groups, necessitating further research. Third, because our study was based on cross-sectional data and we aimed to explore the potential association between AIP and an increased incidence of arthritis, we did not construct a refined predictive model and only roughly assessed the role of AIP. Thus, the consistency of the results must be validated in more detail via longitudinal studies. Finally, similar to observational epidemiological studies, residual confounding from unmeasured risk factors may lead to some degree of bias, indicating the need for broader research on additional confounding factors in the future.

Conclusion

In summary, this study establishes a significant correlation between the AIP and the incidence of arthritis among US adults. As a novel biomarker, AIP holds promise in enhancing arthritis risk assessment and management strategies. Clinicians should be vigilant when monitoring patients with elevated AIP levels, as this may indicate a high risk of developing arthritis, ultimately contributing to improved prevention strategies and healthcare outcomes. Future research should aim to elucidate the underlying mechanisms linking AIP to arthritis and explore the potential of AIP in clinical settings to refine risk stratification and treatment approaches.

Supplemental Material

sj-pdf-1-imr-10.1177_03000605251383294 - Supplemental material for Relationship between the atherogenic index of plasma and arthritis in American adults: Findings from the National Health and Nutrition Examination Survey 2005–2018

Supplemental material, sj-pdf-1-imr-10.1177_03000605251383294 for Relationship between the atherogenic index of plasma and arthritis in American adults: Findings from the National Health and Nutrition Examination Survey 2005–2018 by Yingjie Ge, Lizhi Zhang, Yuming Luo, Lin Yang, Ting He, Yueqi Wang, Xiumin Ou, Jianping Zhang, Zhenquan Sun and Jie Xie in Journal of International Medical Research

Footnotes

Acknowledgements

Not applicable.

Author contributions

Yingjie Ge and Jie Xie conceived and designed the project. Lin Yang, Yuming Luo, Ting He, Yueqi Wang, Jianping Zhang, Zhenquan Sun, and Xiumin Ou collected and analyzed the data. Yingjie Ge and Lizhi Zhang drafted the manuscript. All authors have reviewed and approved the final manuscript.

Consent for publication

All authors have provided consent to the publication of this study.

Data availability statement

The datasets generated during the current study are available in the database ().

Declaration of conflicting interests

The authors have no relevant financial or nonfinancial interests to disclose.

Ethical consideration

The NHANES received ethical approval from the NCHS Institutional Review Board, and all participants provided informed consent.

Funding

This work was supported by the Guangzhou Chinese Medicine and Integrated Traditional Chinese, the Western Medicine Science and Technology Project (003214679029), and the Guangzhou Liwan District Science and Technology Plan Project (202201008 and 20230706).

Supplemental material

Supplemental material for this article is available online.

References

Senthelal

Ardeshirzadeh

, et al. Arthritis. In: StatPearls. Treasure Island (FL): StatPearls, 2025.

Tang

CH.

Research of pathogenesis and novel therapeutics in arthritis 2.0. Int J Mol Sci 2020; 21: 8125.

Lawrence

Felson

Helmick

; National Arthritis Data Workgroupet al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part II. Arthritis Rheum 2008; 58: 26–35.

GBD 2017 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2018; 392: 1789–1858.

Sowers

Karvonen-Gutierrez

CA.

The evolving role of obesity in knee osteoarthritis. Curr Opin Rheumatol 2010; 22: 533–537.

Cross

Smith

Hoy

, et al. The global burden of hip and knee osteoarthritis: estimates from the global burden of disease 2010 study. Ann Rheum Dis 2014; 73: 1323–1330.

Puig-Junoy

Ruiz Zamora

Socio-economic costs of osteoarthritis: a systematic review of cost-of-illness studies. Semin Arthritis Rheum 2015; 44: 531–541.

Sokka

Kautiainen

Pincus

; QUEST-RAet al. Work disability remains a major problem in rheumatoid arthritis in the 2000s: data from 32 countries in the QUEST-RA study. Arthritis Res Ther 2010; 12: R42.

Fernandez-Macias

Ochoa-Martinez

Varela-Silva

, et al. Atherogenic index of plasma: novel predictive biomarker for cardiovascular illnesses. Arch Med Res 2019; 50: 285–294.

10.

Dobiasova

Frohlich

The plasma parameter log (TG/HDL-C) as an atherogenic index: correlation with lipoprotein particle size and esterification rate in apoB-lipoprotein-depleted plasma (FER(HDL)). Clin Biochem 2001; 34: 583–588.

11.

Zhang

Dai

, et al. High-density lipoprotein cholesterol and apolipoprotein A1 in synovial fluid: potential predictors of disease severity of primary knee osteoarthritis. Cartilage 2021; 13: 1465S–1473S.

12.

Zhou

Guo

Wang

, et al. The cross-sectional and longitudinal effect of hyperlipidemia on knee osteoarthritis: results from the Dongfeng-Tongji cohort in China. Sci Rep 2017; 7: 9739.

13.

Maddah

Mahdizadeh

Association of metabolic syndrome and its components with knee osteoarthritis. Acta Med Iran 2015; 53: 743–748.

14.

Rodríguez-Carrio

Alperi-López

López

, et al. High triglycerides and low high-density lipoprotein cholesterol lipid profile in rheumatoid arthritis: a potential link among inflammation, oxidative status, and dysfunctional high-density lipoprotein. J Clin Lipidol 2017; 11: 1043–1054.e2.

15.

Hao

Zhuang

Wang

MJ.

The association between the atherogenic index of plasma (AIP) and sarcopenia in adults aged 20-59: insight from NHANES (2011-2018). Asian J Surg 2024; S1015–9584(24)02147-X.

16.

Hernandez

Olmos

Pariente

, et al. The atherogenic index of plasma is related to a degraded bone microarchitecture assessed by the trabecular bone score in postmenopausal women: The Camargo Cohort Study. Maturitas 2021; 148: 1–6.

17.

March

Schwarz

Carfrae

, et al. Clinical validation of self-reported osteoarthritis. Osteoarthritis Cartilage 1998; 6: 87–93.

18.

Fan

Cai

, et al. Association of urinary phthalate metabolites with osteoarthritis in American adults: results from the national health and nutrition examination survey 2003-2014. Chemosphere 2021; 268: 128807.

19.

Chang

Chiang

EPI

Chang

, et al. The sizes and composition of HDL-cholesterol are significantly associated with inflammation in rheumatoid arthritis patients. Int J Mol Sci 2023; 24: 10645.

20.

Min

Yao

, et al. Association between atherogenic index of plasma control level and incident cardiovascular disease in middle-aged and elderly Chinese individuals with abnormal glucose metabolism. Cardiovasc Diabetol 2024; 23: 54.

21.

Tien

Wang

Lee

, et al. Comparative predictive efficacy of atherogenic indices on metabolic syndrome in patients with schizophrenia. Schizophr Res 2023; 262: 95–101.

22.

Xiao

Wang

Zhang

, et al. Association of systemic immune inflammation index with estimated pulse wave velocity, atherogenic index of plasma, triglyceride-glucose index, and cardiovascular disease: a large cross-sectional study. Mediators Inflamm 2023; 2023: 1966680.

23.

Alifu

Xiang

Zhang

, et al. Association between the atherogenic index of plasma and adverse long-term prognosis in patients diagnosed with chronic coronary syndrome. Cardiovasc Diabetol 2023; 22: 255.

24.

Zhang

Hou

Guo

, et al. Lipid peroxidation in osteoarthritis: focusing on 4-hydroxynonenal, malondialdehyde, and ferroptosis. Cell Death Discov 2023; 9: 320.

25.

Shen

Wang

Cheng

, et al. Pterostilbene alleviated NAFLD via AMPK/mTOR signaling pathways and autophagy by promoting Nrf2. Phytomedicine 2023; 109: 154561.

26.

Yao

Sun

, et al. Chondrocyte ferroptosis contribute to the progression of osteoarthritis. J Orthop Translat 2021; 27: 33–43.

27.

Messner

Bernhard

Smoking and cardiovascular disease: mechanisms of endothelial dysfunction and early atherogenesis. Arterioscler Thromb Vasc Biol 2014; 34: 509–515.

28.

Gianfrancesco

Crowson

CS.

Where there's smoke, there's a joint: passive smoking and rheumatoid arthritis. Arthritis Rheumatol 2021; 73: 2161–2162.

29.

Chang

Yang

Kim

, et al. Smoking and rheumatoid arthritis. Int J Mol Sci 2014; 15: 22279–22295.

30.

Stea

Wandel

Mansoor

, et al. BMI, lipid profile, physical fitness and smoking habits of young male adults and the association with parental education. Eur J Public Health 2009; 19: 46–51.

31.

Lara

Amigo

Association between education and blood lipid levels as income increases over a decade: a cohort study. BMC Public Health 2018; 18: 286.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB