Association between novel inflammatory markers and non-alcoholic fatty liver disease: a cross-sectional study

Abstract

Objective

This study aimed to investigate the association between novel inflammatory markers (NIMs) and non-alcoholic fatty liver disease (NAFLD).

Methods

A total of 6306 subjects were enrolled in this cross-sectional study. NIMs, including neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), C-reactive protein to albumin ratio (CAR), lymphocyte to monocyte ratio (LMR), systemic immune-inflammation index (SII) and prognostic nutritional index (PNI), were calculated. The prevalence of NAFLD and its association with NIMs were assessed by multivariable logistic regression analysis. Subgroup analysis were performed based on age, sex and BMI.

Results

The prevalence of NAFLD was 52.5% in the study population. Compared with non-NAFLD subjects, NAFLD patients were older and more frequent in females. The prevalence of NAFLD progressively increased among the higher quartile groups of CAR, LMR, SII and PNI (P-trend < 0.05), whereas it progressively decreased among the higher quartile group of NLR and PLR (P-trend < 0.05). According to multivariable logistic regression analysis, the highest quartile (Q4) had a significantly higher risk of NAFLD compared with Q1 in LMR [odds ratio (OR): 1.43; 95% confidence interval (CI): 1.17–1.75; P-trend < 0.001] and PNI (OR: 1.92; 95% CI: 1.57–2.35; P-trend < 0.001). The subgroup analysis showed a stronger association of PNI with NAFLD.

Conclusion

The study highlights the association between NIMs and NAFLD, with LMR and PNI identified as potential non-invasive markers of inflammation in NAFLD. Specifically, PNI exhibited the strongest association and may serve as a valuable marker for assessing inflammation in NAFLD.

Introduction

Non-alcoholic fatty liver disease (NAFLD) is a condition closely associated with obesity, type 2 diabetes mellitus (T2DM), hypertension, hyperlipidaemia and metabolic syndrome. NAFLD is a significant cause of end-stage liver disease [1], primary liver cancer and liver transplantation [2] and is currently the fastest-growing cause of liver-related deaths worldwide [3]. Globally, the prevalence of NAFLD is estimated to be around 25% [4], with up to 80% of obese patients and 47.3–63.7% of patients with T2DM suffering from the condition [5,6]. As a result, NAFLD poses a significant economic burden on healthcare systems. In order to better understand the factors and mechanisms underlying NAFLD, more comprehensive research is necessary.

Histologically, NAFLD encompasses a range of liver diseases from hepatic steatosis to non-alcoholic steatohepatitis characterised by lobular inflammation and hepatocellular ballooning (with or without fibrosis) to more advanced stages including cirrhosis and hepatocellular carcinoma [7]. To date, there is no reliable biomarker that can be used to accurately diagnose and stage NAFLD [8,9]. It is now well established that chronic low-grade inflammation plays a key role in the development of NAFLD, with a complex pathophysiological mechanism that is regulated by a variety of intrahepatic and extrahepatic factors [10]. Studies reported excellent performance of inflammatory markers in the diagnosis and prognosis of cancer, cardiovascular disease and infectious diseases [11–13]. However, evidence comparing the correlation of multiple novel inflammatory markers (NIMs) with NAFLD is lacking.

In this study, we conducted a large cross-sectional investigation at a community hospital in Beijing, China to explore the association between multiple NIMs and NAFLD, and to identify optimal non-invasive markers of inflammation in NAFLD.

Methods

Study design and population

Tongzhou Cohort Study (ClinicalTrials.gov Identifier: NCT05156580) is a cohort study conducted by Beijing Friendship Hospital, Capital Medical University. General information of subjects undergoing routine physical examination in Yongshun Community, Tongzhou District, Beijing is collected annually to investigate the health status and prognosis of the general population. This cross-sectional study analysed 6306 residents aged ≥18 years who underwent annual health check-ups at the centre from June to December 2021. Exclusion criteria included missing data and daily alcohol consumption. The study complied with the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Ethics Committee of Beijing Friendship Hospital, Capital Medical University (Beijing, China) (approval no. 2021-P2-163-02). All subjects provided written informed consent.

Data collection and measurements

All subjects were interviewed by trained physicians using a predesigned questionnaire to collect information on age, gender, alcohol consumption, smoking, sleeping, exercise, marital status, residential status, education level and medical history. Information on alcohol consumption was collected in three areas: drinking history, type of alcoholic beverage and approximate amount of alcohol consumed per occasion. Smoking was recorded as current smoker, ex-smoker or non-smoker. Hours of sleep per day were self-reported by the subjects. Regular exercise was defined as more than 30 min per session and at least two times a week. Marital status included single, married, divorced and widowed. Residential status was classified as living alone or cohabiting. Education level was defined in two classifications: below middle school or above high school. Comorbidities such as hypertension and diabetes were self-reported by subjects. Weight and height (without outdoor clothing and shoes) were measured standing on a calibrated scale. BMI was calculated as weight in kilograms divided by height in metres squared. Waist circumference (WC) was measured at the umbilical level using a non-retractable tape measure without pressure on the body surface. Central obesity was defined as WC ≥ 90 cm in men and ≥85 cm in women. Venous blood was collected after an overnight fast of >8 h to measure complete blood count and biochemical parameters using automated biochemistry analysers (7180) from Hitachi.

Assessment of non-alcoholic fatty liver disease

The diagnosis of NAFLD was based on the finding of hepatic steatosis on liver ultrasound, which excluded acute or chronic liver disease or secondary steatosis due to excessive alcohol or drug use [14]. Given the different types and degrees of alcoholic beverages, people with daily drinking habits were excluded from this study to avoid the effects of alcohol. Liver ultrasound was performed by experienced sonographers who were unaware of the study protocol. Fatty liver was defined by at least two of the following three findings: (1) diffusely increased liver near-field ultrasound echo (‘bright liver’), (2) liver echo greater than kidney and (3) vascular blurring and the gradual attenuation of a far-field ultrasound echo [15].

Calculation of novel inflammatory markers

Neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), C-reactive protein (CRP) to albumin ratio (CAR), lymphocyte to monocyte ratio (LMR), systemic immune-inflammation index (SII) and prognostic nutritional index (PNI) were used as NIMs to assess the correlation with NAFLD. NLR was calculated as neutrophil count (10⁹/l) divided by lymphocyte count (10⁹/l). PLR was calculated as platelet count (10⁹/l) divided by lymphocyte count (10⁹/l). CAR was calculated as CRP count (mg/l) divided by albumin count (g/l). LMR was calculated as lymphocyte count (10⁹/l) divided by monocyte count (10⁹/l). SII was calculated as platelet count (10⁹/l) × neutrophil count (10⁹/l)/lymphocyte count (10⁹/l) [16]. PNI was calculated as albumin count (g/l) + 5 × lymphocyte count (10⁹/l) [17].

Statistical analysis

Kolmogorov–Smirnov test and histogram were combined to judge the normality of continuous variables. Continuous variables were expressed as the mean ± SD or median with interquartile range, and categorical variables were expressed as numbers with percentages. Comparisons of characteristics by NAFLD status (with/without) were assessed using Student’s t-tests or Mann–Whitney U test for continuous variables and Chi-square test for categorical variables. Z-transform [formula: Z_i = (X_i –x)/s] was used to standardise NIMs to make them comparable. To obtain a deeper understanding of the relationship between NIMs levels and the prevalence of NAFLD, we next divided the study population into four groups according to every NIM quartile: NLR (Q1: <1.34, Q2 : 1.34 to <1.73, Q3 : 1.73 to <2.21, Q4: ≥2.21), PLR (Q1: <96.24, Q2 : 96.24 to <119.02, Q3 : 119.02 to <146.53, Q4: ≥146.53), CAR (Q1: <0.07, Q2 : 0.07 to <0.11, Q3 : 0.11 to <0.13, Q4: ≥0.13), LMR (Q1: <4.70, Q2 : 4.7 to <5.85, Q3 : 5.85 to <7.21, Q4: ≥7.21), SII (Q1: <298.00, Q2 : 298.00 to <401.61, Q3 : 401.61 to <540.87, Q4: ≥540.87), PNI (Q1: <51.85, Q2 : 51.85 to <54.20, Q3 : 54.20 to <56.85, Q4: ≥56.85). Adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated using multivariable logistic regression analysis to determine the risk of NAFLD in each NIM quartile and using the lowest quartile as the reference. Multivariable logistic regression was performed using Models 1, 2 and 3 respectively to adjust for the effect of known confounders on the prevalence of NAFLD. Integrating professional knowledge and the results of univariate analysis to consider the variables that need to be included in the model. Model 1 adjusted for sex, age, WC and BMI. Model 2 further adjusted for education, marriage, residential status, smoking, sleep duration and exercise. Model 3 further adjusted for hypertension, diabetes, uric acid, alanine transaminase (ALT), triglyceride (TG), total cholesterol (TC) and high-density lipoprotein cholesterol (HDL-C). To assess the sensitivity of the results, the total population was further divided into age subgroups (<60 and ≥60, year), sex subgroups (male and female) and BMI subgroups (<24, 24–27.9 and ≥28, kg/m²). Multivariable logistic regression was performed for each subgroup using Model 3. All analyses were performed using IBM SPSS (version 25.0) and GraphPad Prism (version 8.0). Two-tailed P values of <0.05 were considered statistically significant.

Results

Prevalence of non-alcoholic fatty liver disease in the general population

A total of 5013 subjects with abdominal ultrasonography examination were included and divided into two groups, consisting of 2631 (52.5%) with NAFLD and 2382 (47.5%) without NAFLD (Fig. 1). The prevalence of NAFLD is higher in the elderly (≥60, 54.2% vs. <60, 49.8%, P = 0.002) and in females (55.6% vs. 40.0%, P < 0.001). The higher the BMI, the higher the prevalence of NAFLD (<24, 19.0% vs. 24–27.9, 50.1% vs. ≥28 kg/m², 79.1%, P < 0.001). Patients with central obesity (69.2% vs. 31.1%), hypertension (61.4% vs. 46.3%) and diabetes (60.5% vs. 50.7%) are more likely to have NAFLD (all P < 0.001) (Table 1).

Fig. 1.

Flowchart of the study population. NAFLD: non-alcoholic fatty liver disease.

Table 1.

Prevalence of non-alcoholic fatty liver disease based on different characteristics

		N	NAFLD	Statistics	P value
Age, year	<60	1992	993 (49.8)	9.197	0.002
	≥60	3021	1638 (54.2)
Sex	Male	993	397 (40.0)	77.630	<0.001
	Female	4020	2234 (55.6)
BMI, kg/m2	<24	1142	217 (19.0)	983.714	<0.001
	24–27.9	2228	1117 (50.1)
	≥28	1640	1297 (79.1)
Central obesity	Yes	2817	1949 (69.2)	718.046	<0.001
	No	2194	682 (31.1)
Hypertension	Yes	2049	1258 (61.4)	110.375	<0.001
	No	2964	1373 (46.3)
Diabetes	Yes	922	558 (60.5)	29.264	<0.001
	No	4091	2073 (50.7)

BMI, body mass index; NAFLD, non-alcoholic fatty liver disease.

Characteristics of the study subjects

Demographic, clinical and laboratory characteristics between individuals with and without NAFLD are described in Table 2. Compared with non-NAFLD subjects, NAFLD patients were older (63.14 vs. 62.47, P = 0.002), had higher BMI, WC, sleep duration, uric acid, platelet count, neutrophil, lymphocyte, monocyte, CRP, ALT, AST, albumin, TG, TC, low-density lipoprotein cholesterol and lower HDL-C. Meanwhile, patients with NAFLD also more frequently presented with female (84.9% vs. 75.0%, P < 0.001), low education, less exercise, diabetes and hypertension. For NIMs, NAFLD patients had higher CAR, LMR, SII, PNI and lower NLR and PLR (all P < 0.05).

Table 2.

Characteristics of the study subjects with and without non-alcoholic fatty liver disease

Variables	Non-NAFLD	NAFLD	Statistics	P value
	n = 2382	n = 2631
Age, year	62.47 ± 7.94	63.14 ± 7.41	−3.117	0.002
Female, n (%)	1786 (75.0)	2234 (84.9)	77.630	<0.001
BMI, kg/m²	24.98 ± 3.03	28.27 ± 3.36	−36.552	<0.001
WC, cm	84.24 ± 8.73	92.00 ± 8.45	−31.938	<0.001
Low education, n (%)	1854 (77.8)	2180 (82.9)	20.083	<0.001
Cohabitation, n (%)	2155 (90.5)	2343 (89.1)	2.722	0.099
Exercise, n (%)	756 (31.7)	696 (26.5)	16.966	<0.001
Sleep duration, h	7.30 ± 1.24	7.40 ± 1.21	−2.911	0.004
History of smoking, n (%)			52.745	<0.001
Current smoker	308 (12.9)	238 (9.0)
Ex-smoker	227 (9.5)	147 (5.6)
Non-smoker	1847 (77.5)	2246 (85.4)
Marriage status, n (%)			1.879	0.604
Single	3 (0.1)	3 (0.1)
Married	2167 (91)	2388 (90.8)
Divorced	30 (1.3)	45 (1.7)
Widowed	182 (7.6)	195 (7.4)
Diabetes, n (%)	364 (15.3)	558 (21.2)	29.264	<0.001
Hypertension, n (%)	791 (33.2)	1258 (47.8)	110.375	<0.001
eGFR, ml/min/1.73 m²	92.3 ± 19.0	92.5 ± 20.0	−0.289	0.773
Uric acid, μmol/l	306.9 ± 76.5	340.3 ± 83.5	−14.798	<0.001
Platelet, 109/l	232.1 ± 57.5	245.1 ± 56.6	−8.085	<0.001
Neutrophil, 109/l	3.43 ± 1.15	3.68 ± 1.12	−7.758	<0.001
Lymphocyte, 109/l	1.93 ± 0.58	2.14 ± 0.62	−12.072	<0.001
Monocyte, 109/l	0.35 ± 0.11	0.36 ± 0.10	−4.997	<0.001
CRP, mg/l	4.7 (2.8, 5.3)	4.9 (3.2, 5.8)	−8.072	<0.001
ALT, U/l	16.4 (12.7, 21.0)	20.5 (15.8, 29.0)	−20.181	<0.001
AST, U/l	20.0 (17.4, 23.1)	21.0 (17.9, 25.1)	−7.883	<0.001
TBIL, μmol/l	11.48 ± 4.83	11.32 ± 4.63	1.168	0.243
Albumin, g/l	44.08 ± 2.40	44.41 ± 2.40	−4.819	<0.001
TG, mmol/l	1.32 (1.01, 1.75)	1.80 (1.36, 2.42)	−24.138	<0.001
TC, mmol/l	5.54 ± 1.18	5.66 ± 1.19	−3.607	<0.001
LDL-C, mmol/l	3.06 ± 0.88	3.23 ± 0.93	−6.763	<0.001
HDL-C, mmol/l	1.53 ± 0.36	1.36 ± 0.28	18.566	<0.001
Novel inflammatory markers
NLR	1.91 ± 0.87	1.84 ± 0.72	3.106	0.002
PLR	128.1 ± 42.7	122.1 ± 39.0	5.177	<0.001
CAR	0.10 (0.06, 0.12)	0.11 (0.07, 0.13)	−6.724	<0.001
LMR	5.92 ± 1.94	6.23 ± 1.96	−5.458	<0.001
SII	395.2 (292.1, 535.3)	407.0 (302.6, 544.6)	−2.496	0.013
PNI	53.74 ± 3.79	55.09 ± 3.93	−12.355	<0.001

Continuous variables are expressed as mean ± SD or median with interquartile range. Categorical variables are expressed as n (%).

ALT, alanine transaminase; AST, aspartate transaminase; BMI, body mass index; CAR, CRP to albumin ratio; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; LMR, lymphocyte to monocyte ratio; NAFLD, non-alcoholic fatty liver disease; NLR, neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; PNI, prognostic nutritional index; SII, systemic immune-inflammation index; TBIL, total bilirubin; TC, total cholesterol; TG, triglyceride; WC, waist circumference.

Association between novel inflammatory markers and non-alcoholic fatty liver disease

To better examine the relationship between NIMs and the prevalence of NAFLD, we divided the subjects into four groups according to the NIMs quartiles. The prevalence of NAFLD was statistically different in all standardised NIMs quartiles (P < 0.05) except the SII quartile groups (P = 0.147). After trend testing, the prevalence of NAFLD progressively increased in the higher CAR, LMR, SII and PNI quartile groups (P-trend < 0.05), whereas it progressively decreased in the higher NLR and PLR quartile groups (P-trend < 0.05) (Supplementary Table 1, Supplemental digital content 2, http://links.lww.com/EJGH/A952).

Table 3 shows the association between quartiles of every NIM and NAFLD. The highest quartile of NLR (Q4) showed a decreased risk of NAFLD compared with the reference quartile (Q1) (OR: 0.82; 95% CI: 0.68–0.99; P-trend = 0.054) in Model 1. This association persisted in Model 2 (OR: 0.83; 95% CI: 0.69–1.00; P-trend = 0.070) and became statistically significant in Model 3 (OR: 0.78; 95% CI: 0.64–0.96; P-trend = 0.033). In Model 1, the third quartile of PLR (Q3) was associated with a significantly reduced risk of NAFLD compared with Q1 (OR: 0.81; 95% CI: 0.68–0.98; P-trend = 0.050). A similar trend was observed in Model 2 (OR: 0.81; 95% CI: 0.67–0.97; P-trend = 0.041), but the association was attenuated and no longer significant in Model 3 (P-trend = 0.494). LMR exhibited a consistent positive association with NAFLD risk across all models. The highest quartile (Q4) had a significantly higher risk of NAFLD compared with Q1 in all three models (Model 1: OR: 1.44; 95% CI: 1.19–1.75; Model 2: OR: 1.43; 95% CI: 1.18–1.73; Model 3: OR: 1.43; 95% CI: 1.17–1.75, all P-trend < 0.001). PNI demonstrated a consistently strong positive association with NAFLD risk across all models. The highest quartile (Q4) had a significantly higher risk of NAFLD compared with Q1 (Model 1: OR: 2.65; 95% CI: 2.19–3.20; Model 2: OR: 2.65; 95% CI: 2.20–3.21; Model 3: OR: 1.92; 95% CI: 1.57–2.35, all P-trend < 0.001). However, no significant associations were observed between both CAR and SII quartiles and NAFLD risk in all three models. In order to establish comparability among the various NIMs, the Z-transformed values of each NIM were incorporated into Model 3 as continuous variables. The findings revealed a positive correlation between LMR (OR: 1.15; 95% CI: 1.07–1.24; P < 0.001) and PNI (OR: 1.31; 95% CI: 1.22–1.41; P < 0.001) with NAFLD, while the correlations between the remaining NIMs and NAFLD were found to be statistically insignificant (Fig. 2).

Table 3.

Odds ratios and 95% confidence intervals for non-alcoholic fatty liver disease according to quartiles of novel inflammatory markers

	Q1	Q2	Q3	Q4	P-trend
NLR
Model 1	1 (Ref)	0.91 (0.76, 1.09)	0.92 (0.77, 1.10)	0.82 (0.68, 0.99)*	0.054
Model 2	1 (Ref)	0.90 (0.75, 1.09)	0.92 (0.77, 1.11)	0.83 (0.69, 1.00)*	0.070
Model 3	1 (Ref)	0.90 (0.74, 1.09)	0.93 (0.77, 1.13)	0.78 (0.64, 0.96)*	0.033
PLR
Model 1	1 (Ref)	0.91 (0.76, 1.10)	0.81 (0.68, 0.98)*	0.86 (0.71, 1.03)	0.050
Model 2	1 (Ref)	0.93 (0.77, 1.12)	0.81 (0.67, 0.97)*	0.85 (0.71, 1.03)	0.041
Model 3	1 (Ref)	0.96 (0.79, 1.16)	0.90 (0.74, 1.09)	0.95 (0.78, 1.15)	0.494
CAR
Model 1	1 (Ref)	0.94 (0.79, 1.13)	1.13 (0.94, 1.36)	1.17 (0.97, 1.41)	0.032
Model 2	1 (Ref)	0.95 (0.79, 1.14)	1.13 (0.94, 1.36)	1.16 (0.97, 1.40)	0.036
Model 3	1 (Ref)	0.95 (0.78, 1.15)	1.08 (0.89, 1.31)	0.95 (0.78, 1.15)	0.947
LMR
Model 1	1 (Ref)	1.06 (0.88, 1.28)	1.24 (1.03, 1.49)*	1.44 (1.19, 1.75)*	<0.001
Model 2	1 (Ref)	1.07 (0.89, 1.29)	1.25 (1.03, 1.51)*	1.43 (1.18, 1.73)*	<0.001
Model 3	1 (Ref)	1.08 (0.89, 1.31)	1.21 (0.99, 1.48)	1.43 (1.17, 1.75)*	<0.001
SII
Model 1	1 (Ref)	1.05 (0.88, 1.26)	1.15 (0.96, 1.38)	1.14 (0.95, 1.36)	0.103
Model 2	1 (Ref)	1.05 (0.88, 1.26)	1.15 (0.96, 1.38)	1.15 (0.95, 1.38)	0.089
Model 3	1 (Ref)	1.03 (0.85, 1.25)	1.12 (0.92, 1.36)	1.05 (0.87, 1.28)	0.447
PNI
Model 1	1 (Ref)	1.27 (1.06, 1.53)*	1.76 (1.46, 2.11)*	2.65 (2.19, 3.20)*	<0.001
Model 2	1 (Ref)	1.27 (1.05, 1.53)*	1.77 (1.47, 2.13)*	2.65 (2.20, 3.21)*	<0.001
Model 3	1 (Ref)	1.16 (0.95, 1.41)	1.50 (1.24, 1.83)*	1.92 (1.57, 2.35)*	<0.001

Model 1: adjusted for sex, age, WC and BMI. Model 2: further adjusted for education, marriage, residential status, smoking, sleep duration and exercise. Model 3: further adjusted for hypertension, diabetes, uric acid, ALT, TG, TC and HDL-C.

CAR, CRP to albumin ratio; LMR, lymphocyte to monocyte ratio; NAFLD, non-alcoholic fatty liver disease; NIM, novel inflammatory markers; NLR, neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; PNI, prognostic nutritional index; SII, systemic immune-inflammation index.

*P < 0.05.

Fig. 2.

Adjusted odds ratios and 95% confidence intervals of NIMs per 1-SD increase. Model 3 was used: adjusted for sex, age, WC, BMI, education, marriage, residential status, smoking, sleep duration, exercise, hypertension, diabetes, uric acid, ALT, TG, TC and HDL-C. CAR, CRP to albumin ratio; LMR, lymphocyte to monocyte ratio; NAFLD, non-alcoholic fatty liver disease; NIMs, novel inflammatory marker; NLR, neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; PNI, prognostic nutritional index; SII, systemic immune-inflammation index.

Subgroup analysis

The subgroup analysis revealed associations between NIMs and NAFLD based on age, sex and BMI in Model 3 (Supplementary Tables 2–4, Supplemental digital content 3, http://links.lww.com/EJGH/A953). The findings indicated that higher levels of LMR and PNI may be associated with an increased risk of NAFLD. The association of progressively increasing risk of NAFLD with increasing PNI quartiles persisted across age, sex and BMI subgroups (P-trend < 0.001). The positive association between LMR and NAFLD was only present in those aged ≥60 years (P-trend = 0.001), females (P-trend = 0.001) or BMI ≥ 24 kg/m² (P-trend < 0.05). Besides, a positive association between CAR and NAFLD was found in males (P-trend = 0.015) and a negative association between NLR and NAFLD was found in females (P-trend = 0.042).

Discussion

In this cross-sectional study, we investigated the association between NIMs and NAFLD. We mainly found that both LMR and PNI were positively associated with NAFLD. Importantly, after subgroup analysis, PNI remained positively associated with NAFLD, regardless of age, sex and BMI, while the association between LMR and NAFLD was mainly seen in elder women who were overweight or obese.

Although NAFLD currently affects around a quarter of the world’s population [4], its prevalence varies across individuals with different characteristics. This study found that people aged ≥60 years were more likely to have NAFLD. Several other studies have also found a positive correlation between age and NAFLD, however, Koehler et al. [18] found a negative association between NAFLD and age among the elderly population. This suggests an inverted U-shaped relationship between the prevalence of NAFLD and age. Studies have indicated that men are more likely to develop NAFLD in the general adult population. Interestingly, in postmenopausal women, the prevalence of NAFLD is higher, which indicates a potential protective effect of oestrogen [19]. In our study, the mean age of the population was 62.8, and over 98% of the subjects were over 50 years old, which may explain the higher incidence of NAFLD in women. As a hepatic manifestation of the metabolic syndrome, NAFLD is closely associated with obesity, T2DM and hypertension [20], which our study supported.

The hepatic inflammatory response is a crucial factor driving the progression of NAFLD, leading to progressive liver fibrosis and, ultimately, the development of cirrhosis [21]. Various studies in both human and animal models have demonstrated the integral role of immune cells in the pathogenesis of NAFLD [22]. This relationship may be attributed to free fatty acids leading to lipid peroxidation and cytokine production [23]. Previous studies have shown positive associations between certain NIMs and NAFLD, but have not adequately adjusted for confounders and have shown inconsistent results in subgroups [16,17,24]. While Xie et al. [16] found a minor association between elevated SII levels and hepatic steatosis, our study reported a negative association between SII and NAFLD. Differences in ethnicity, statistical methods and the definition of NAFLD may account for these variations. The most striking finding of this study is the strongest association between PNI and NAFLD, as evidenced by two observations: (1) the prevalence of NAFLD increased steadily with rising PNI quartiles and (2) the positive association between PNI and NAFLD was the most significant in the whole population and persisted across subgroups of the population. Our results complement those of Pozza et al. [17], who found that PNI and PLR were associated with biopsy-diagnosed liver fibrosis in obese patients. Regarding the pathophysiological mechanisms, lymphocytes are recruited to the liver in response to inflammation caused by fat accumulation. Within the liver, lymphocytes differentiate into different subtypes, including T cells, B cells and natural killer cells. Subtypes of T cells may release pro-inflammatory cytokines that trigger liver damage and subsequent disease progression [25]. Additionally, B lymphocytes can secrete cytokines and antibodies, augmenting inflammation and liver damage [26]. Therefore, we postulated that PNI, a lymphocyte-derived marker, holds good predictive value in all stages of NAFLD and can be used for managing NAFLD. Furthermore, NIM has also exhibited significant associations with other metabolic disorders. The results of a meta-analysis including 16 studies showed that pregnant women with gestational diabetes mellitus (GDM) had significantly higher levels of NLR compared with those without GDM [27]. The presence of an elevated NLR has also been observed to be associated with suboptimal glycaemic control in patients with DM [28]. The findings highlight the importance of recognising inflammation’s pivotal role in metabolic disorders.

Upon conducting subgroup analysis, this study has yielded some intriguing findings regarding the correlation between NIMs and NAFLD among elder women and those with a higher BMI. These results may reflect the specific immunological characteristics of this particular population. Females generally exhibit greater immune responses than males [29], and sex steroids can directly affect immune function. Furthermore, sex hormone depletion and increased inflammation have been linked to metabolic syndrome, cardiovascular disease and osteoporosis in postmenopausal women [30]. As we age, the expansion of visceral fat becomes increasingly prevalent. Insulin resistance and lipid dysregulation are two factors that fuel inflammation in both metabolic and non-metabolic diseases linked to ageing and obesity [30]. Although liver biopsy showed a similar rate of lobular inflammation in lean NAFLD patients compared with obese patients [31], our results suggested that the systemic inflammatory status of lean NAFLD appeared to be lower. The mechanisms involved are likely to be complex and further research is warranted.

There are some limitations to this study. First, given that this is a cross-sectional study, the causal relationship between NIMs and NAFLD cannot be assessed. Second, although we have taken into account the influence of socio-demographics and lifestyle habits on NAFLD compared with previous studies [16,17,24], there may still be some potential confounding factors. Third, there was a lack of information on viral hepatitis and malignancy, which may affect the exclusion of subjects. Fourth, this study mainly included people aged ≥50 years, thus results may be different in younger populations. Fifth, NAFLD was diagnosed by abdominal ultrasonography, which is less sensitive when steatosis is mild or when it progresses to steatohepatitis and liver fibrosis. However, although liver biopsy is the gold standard for diagnosing NAFLD, the cost of invasive procedures and the risk of potential complications make it unsuitable for screening purposes [32].

NAFLD did exhibit a higher inflammatory status than non-NAFLD. LMR and PNI were positively associated with NAFLD. Subgroup analysis revealed that the above association between NIMs and NAFLD was mainly present in elder women and the overweight obese, who may have a more significant inflammatory state. Of note, PNI exhibited the strongest association and may serve as a valuable marker for assessing inflammation in NAFLD. Further validation studies are needed to confirm these findings and explore the clinical utility of PNI in NAFLD management.

Acknowledgements

The statistical analysis in this study was expertly guided by Shanshan Wu, for which we express our gratitude.

This study was financially supported by grants from the Science and Technology Plan of Beijing Tongzhou (Beijing, China) (grant no. KJ2022CX036) and the Summit Talent Plan of the Beijing Hospital Management Center (Beijing, China) (grant no: DFL20190101).

The datasets generated during and/or analysed during the current study are not publicly available, but are available from the corresponding author on reasonable request.

U.S. National Library of Medicine. ClinicalTrials.gov Identifier: NCT05156580.

Conflicts of interest

There are no conflicts of interest.