The association of red blood cell markers with NAFLD and advanced liver fibrosis

DOI: https://doi.org/10.21203/rs.3.rs-2026166/v1

Abstract

Background: Nonalcoholic fatty liver disease (NAFLD) is an increasing disease related with metabolic syndrome (MetS). The associations between red blood cell (RBC) markers and MetS have been reported. However, whether RBC markers, including RBC count, mean corpuscular hemoglobin concentration(MCHC), and red blood cell distribution width (RDW)are associated with the risk of NAFLD and advanced liver fibrosis are still unclear.

Methods: We conducted a nationally representative cross-sectional study based on National Health and Nutrition Examination Survey (NHANES) 2017-2018. NAFLD was diagnosed when controlled attenuation parameter (CAP) values ≥263 dB/m after exclusion of hepatitis B or C virus infection and significant alcohol intake. Advanced liver fibrosis was confirmed when liver stiffness measurement (LSM)≥8.6kPa. Weighted multivariable logistic regression models were performed to investigate the associations of RBC markers with NAFLD and advanced liver fibrosis.

Results: 3563 participants were finally included. Of all the participants, 1818 (51%) were diagnosed as NAFLD and 288(8%) had advanced liver fibrosis. After fully adjusting potential confounders, RBC count was positively associated with NAFLD (OR=1.5, 95% CI=1.2-1.8, P<0.001) or CAP (β=12.1, 95% CI= 8.1-16.2, p <0.001) while MCHC was also positively associated with NAFLD (OR=1.2, 95% CI=1.1- 1.3, P=0.003) or CAP (β=2.7, 95% CI= 0.6-4.8, p=0.013). After fully adjusting potential confounders, RDW was positively associated with advanced liver fibrosis (OR=1.3, 95% CI=1.1-1.4, P<0.001) and LSM (β=0.2, 95% CI= 0.1-0.3, p =0.005).However, there was no significant difference in RBC count or MCHC with advanced liver fibrosis or LSM. Moreover, no significant difference was detected between RDW and NAFLD or CAP.

Conclusion: We revealed that RBC count and MCHC were positively associated with NAFLD while RDW was positively associated with advanced liver fibrosis in Americans.

Introduction

Due to dramatic changes in lifestyle and dietary pattern in recent years, non-alcoholic fatty liver disease (NAFLD) has been persistently increased to be the most common cause of chronic liver disease globally[1]. The population suffering from NAFLD in America is predicted to increase to 100.9 million in 2030[2]. NAFLD could increase both the hepatic-associated mortality and the risk of cardiovascular diseases, malignant tumors, and metabolic disorders, leading to a significantly higher risk of all-cause mortality[3]. Although the leading causes of death in people with NAFLD are cardiovascular disease and extra-hepatic malignancy, advanced liver fibrosis is a key prognostic marker for liver-related outcomes and overall mortality[4]. NAFLD is prevailing as a public health challenge and causes a considerable economic burden. A convenient and effective risk predictor for NAFLD and advanced liver fibrosis is currently lacking.

The widely recognized “multiple parallel hits” considered NAFLD was closely related to insulin resistance, oxidative stress, and inflammation[5]. As an easily available laboratory index, red blood cell (RBC) is responsible for delivering oxygen, nitric oxide, carbon dioxide, and removing reactive oxygen and nitrogen species under physiological conditions[6]. Recently, RBC is revealed to function in modulating immune-related inflammation responses and oxidative stress[7]. Previous studies have revealed that a higher RBC count was correlated with the higher risk of metabolic syndrome [8, 9]. NAFLD has been well-proved to be related with metabolic syndrome, manifesting an inseparable relationship. Few literatures have studied the relationships between RBC markers and NAFLD and liver fibrosis, including RBC count, mean corpuscular hemoglobin concentration(MCHC) and red blood cell distribution width (RDW), but the conclusions were contradictory[10–13]. There were limited studies that have investigated the association between RBC markers and NAFLD and advanced liver fibrosis in the American general population.

Therefore, we conducted the study from the 2017–2018 National Health and Nutrition Examination Survey (NHANES) to evaluate the association between RBC markers and the risk of NAFLD detected by FibroScan® in the American general population.

Methods

Data Source

NHANES is a large, well-designed, nationally representative population-based cross-sectional survey, conducted by the National Center for Health Statistics of the Center for Disease Control and Prevention in the US[14]. Full detailed information of the design and operation is available at https://www.cdc.gov/nchs/nhanes. All protocols were approved by Research Ethics Review Board of the National Center for Health Statistics of the Center and informed consent was attained from participants.

Of all the 9254 participants in NHANES 2017–2018, we excluded those without available data of mobile examination center (MEC) exam (N = 550), considered ineligible for different reasons(N = 258) including participant refusal, limitations or limited time for vibration control transient elastography(VCTE) or not available (N = 2459), partial VCTE exam(N = 493), with evidence of hepatitis B virus(N = 27) or hepatitis C virus infection(N = 86), with significant alcohol intake(N = 752), and without available RBC data(N = 204) and younger than 20 years(N = 862). Finally, 3563 participants were included in this analysis.

Variables

In this study, the independent variable was RBC count, MCHC and RDW. The dependent variable was the risk of NAFLD and advanced liver fibrosis. Diagnostic criteria for NAFLD: liver steatosis after exclusion of hepatitis B or C virus infection and significant alcohol intake[15]. Participants were regarded as indicative of liver steatosis when controlled attenuation parameter(CAP) ≥ 263 dB/m and advanced liver fibrosis (≥ F3)when liver stiffness measurement (LSM) ≥ 8.6kPa, respectively[16]. Hepatitis C virus infection was confirmed by positive for antibody test or viral RNA and hepatitis B virus infection was defined as positive for surface antigen test[17]. Alcohol intake was assessed with data during the past year. Significant alcohol intake was more than 3 drinks per day for men and more than 2 drinks per day for women. Body mass index (BMI) was calculated as dividing kilograms by weight in meters squared. Type 2 diabetes mellitus was diagnosed according to the criteria in previous literatures[18]. Demographic characteristics and laboratory methods are publicly available at the website (https://wwwn.cdc.gov/nchs/data/nhanes). Family income levels were defined by the poverty income ratio (PIR) as low (PIR < 1.3), middle (PIR 1.3–3.5), and high (PIR > 3.5). Overweight was defined as BMI ≥ 25 kg/m2 in Caucasians or BMI ≥ 23 kg/m2 in Asians. Obesity was defined as BMI ≥ 30 kg/m2. Hypertension was defined based on previous physician diagnosis, taking prescribed medicine to decrease blood pressure (BP), or BP of ≥ 140/90 mmHg. Dyscholesterolemia was defined as serum total cholesterol (TC) > 200 mg/dL, low-density lipoprotein cholesterol (LDL-C) ≥ 130 mg/dL, or high-density lipoprotein cholesterol(HDL-C) < 40 mg/dL for men and < 50 mg/dL for women. Anemia was defined as hemoglobin level of lower than 12g/dL. RBC count on whole blood was measured by the UniCelDxH 800 Analyzer (Beckman Coulter, Inc. Fullerton, CA). For more detailed information, refer to https://wwwn.cdc.gov/nchs/data/nhanes/2017- 2018/labmethods/CBC-J- MET-508.pdf.

Statistical Methods

The sample weights were taken into account as recommended by National Center for Health Statistics. Categorical variables were presented as numbers and weighted proportions. Continuous variables were shown as weighted median (quartile 1- quartile 3) since they were non-normally distributed. Statistical analyses were conducted using package R version 3.4.3 (http://www.R-project.org) and EmpowerStats software. P value < 0.05 was considered statistically significant. After quartile classification of the independent variables, weighted multivariable logistic regression models and linear regression models were conducted to evaluate the correlation between RBC count and NAFLD or CAP, MCHC and NAFLD or CAP, RDW and advanced liver fibrosis or LSM. Three models were created: Model 1: no covariates were adjusted; model 2: age, sex, and race were adjusted; model 3: age, sex, race/ethnicity, smoking, obesity, diabetes, dyscholestrolemia, hypertension, physical activity, and anemia were fully adjusted.

Results

Baseline characteristics of the study participants

Weighted baseline characteristics of the study participants based on NAFLD and advanced liver fibrosis were presented in Table 1.

Table 1

Weighted characteristic of participants according to NAFLD and advanced liver fibrosis.

 

Overall

NAFLD(CAP ≥ 263dB/m)

Advanced liver fibrosis(LSM ≥ 8.6kPa)

 

YES

NO

P value

YES

NO

P value

N(%)

3563(100%)

1818 (51%)

1745(49%)

 

288(8%)

3275(92%)

 

VCTE measurement

             

CAP(dB/m)

264(220–308)

307 (283–339)

219 (195–241)

< 0.001

260 (218–302)

319 (277–358)

< 0.001

LSM(kPa)

4.9 (4.0-6.1)

5.4 (4.3–6.7)

4.5 (3.7–5.5)

< 0.001

11.1(9.6–14.6)

4.8 (4-5.8)

< 0.001

Demographics

             

Age(years)

55 (38–66)

57 (44–67)

51 (33–65)

< 0.001

61 (50–69)

54(38–66)

< 0.001

Men: N(%)

1692 (47.5%)

946(52%)

746 (42.8%)

< 0.001

171 (59.4%)

1521 (46.4%)

< 0.001

Race/Ethnicity: N(%)

     

< 0.001

   

0.023

Mexican American

438 (12.3%)

276 (15.2%)

162 (9.3%)

 

40 (13.9%)

398 (12.2%)

 

Other Hispanic

336 (9.4%)

167 (9.2%)

169 (9.7%)

 

33 (11.5%)

303 (9.3%)

 

Non-Hispanic White

1202(33.7%)

639 (35.1%)

563 (32.3%)

 

106 (36.8%)

1096 (33.5%)

 

Non-Hispanic Black

836(23.5%)

359 (19.7%)

477 (27.3%)

 

63 (21.9%)

773 (23.6%)

 

Non-Hispanic Asian

568 (15.9%)

288 (15.8%)

280 (16.0%)

 

27 (9.4%)

541 (16.5%)

 

Other race

183 (5.1%)

89 (4.9%)

94 (5.4%)

 

19 (6.6%)

164 (5.0%)

 

Education attainment

     

0.171

   

0.043

Less than high school graduate

680 (19.1%)

369 (20.3%)

311 (17.8%)

 

62 (21.5%)

618 (18.9%)

 

High school graduate or GED

806 (22.6%)

406 (22.3%)

400 (22.9%)

 

78 (27.1%)

728 (22.2%)

 

College or above

2077 (58.3%)

1043 (57.4%)

1034 (59.3%)

 

148 (51.4%)

1929 (58.9%)

 

Family income level

     

0.425

   

0.244

Low

811 (22.8%)

407 (22.4%)

404 (23.2%)

 

71 (24.7%)

740 (22.6%)

 

Middle

1722 (48.3%)

898 (49.4%)

824 (47.2%)

 

146 (50.7%)

1576 (48.1%)

 

High

1030 (28.9%)

513 (28.2%)

517 (29.6%)

 

71 (24.7%)

959 (29.3%)

 

Smoking status: N(%)

     

< 0.001

   

< 0.001

Current smoker

479 (13.4%)

206 (11.3%)

273 (15.6%)

 

33 (11.5%)

446 (13.6%)

 

Former smoker

876 (24.6%)

510 (28.1%)

366 (21%)

 

98 (34.0%)

778 (23.8%)

 

Nonsmoker

2208 (62.0%)

1102 (60.6%)

1106 (63.4%)

 

157 (54.5%)

2051 (62.6%)

 

Physical activity:N(%)

     

< 0.001

   

0.154

< 600 MET-minutes/week

1373 (38.5%)

773 (42.5%)

600 (34.4%)

 

126 (43.8%)

1247 (38.1%)

 

600 ~ 7999 MET-minutes/week

1620 (45.5%)

772 (42.5%)

848 (48.6%)

 

122 (42.4%)

1498 (45.7%)

 

≥ 8000 MET-minutes/week

570 (16.0%)

273 (15%)

297 (17%)

 

40 (13.9%)

530 (16.2%)

 

Biochemical laboratory tests

             

TC (mg/dL)

184(161–213)

187(163–216)

183(160–210)

< 0.001

178(152–207)

185(162–213)

< 0.001

HDL-C(mg/dL)

51(43–61)

47(41–56)

55(47–66)

< 0.001

45(40–55)

52(43–62)

< 0.001

Glycohemoglobin(%)

5.6(5.3-6.0)

5.8(5.4–6.4)

5.5 (5.2–5.8)

< 0.001

6.0 (5.5–6.7)

5.6 (5.3-6.0)

< 0.001

AST(IU/L)

19(16–23)

20(16–25)

19(16–22)

< 0.001

21.2 (18–29)

19(16–23)

< 0.001

ALT(IU/L)

18(13–25)

20(15–29)

15(12–21)

< 0.001

22(16–36)

17(13–24)

< 0.001

GGT(IU/L)

21(15–31)

25(18–38)

18(13–26)

< 0.001

32(22–58)

21(14–29)

< 0.001

Serum albumin(g/L)

41(39–43)

41(38–43)

41(39–43)

< 0.001

40(38–42)

41(39–43)

< 0.001

Platelet count(109/L)

234(199–277)

236(201–280)

230(198–272)

0.016

225(182–267)

235(200–277)

< 0.001

Uric acid(mg/dl)

5.3(4.4–6.3)

5.6 (4.8–6.7)

5.0 (4.1–5.9)

< 0.001

0.4 (0.3–0.6)

0.4 (0.3–0.5)

< 0.001

SC(mg/dl)

0.9 (0.7-1.0)

0.9 (0.7-1.0)

0.8 (0.7-1.0)

0.122

0.9 (0.7–1.1)

0.8 (0.7-1.0)

0.003

Total bilirubin(mg/dl)

0.4 (0.3–0.6)

0.4 (0.3–0.5)

0.4 (0.3–0.6)

0.852

0.4 (0.3–0.6)

0.4 (0.3–0.5)

< 0.001

HS-CRP(mg/dl)

2(0.9–4.3)

2.6 (1.2–5.3)

1.4 (0.7–3.5)

< 0.001

3.7 (1.8–7.3)

1.8 (0.9-4.0)

< 0.001

Ferritin(ng/ml)

111(53.1–198)

124(66.8–222)

94.3(44–169)

< 0.001

(74.8-266.5)

107(51.5–193)

< 0.001

Hemoglobin(g/dL)

14(13.1–15)

14.2(13.2–15.1)

13.8 (13-14.8)

< 0.001

14.1(13–15)

14(13.1–15)

0.852

RBC markers

             

RBC count (1012/L)

4.7(4.4-5.0)

4.8 (4.5–5.1)

4.6 (4.3-5.0)

< 0.001

4.8 (4.4–5.1)

4.7 (4.4-5.0) 4.8 (4.4–5.1)

0.221

MCHC(g/dL)

33.6 (33.0-34.1)

33.6(33.1–34.2)

33.5(32.9–34)

< 0.001

33.5(32.9–34.2)

33.6(33.0-34.1)

0.801

RDW (%)

13.6(13.1–14.3)

13.6(13.1–14.4)

13.5(13-14.2)

< 0.001

14(13.4–14.8)

13.5(13.1–14.2)

< 0.001

Metabolic disorders

             

WC(cm)

98(88.8-109.4)

105(96.5-116.6)

91.5(82.2–99)

< 0.001

113.5(100–127)

97.0(88.0-107.5)

< 0.001

BMI(kg/m2)

28.3(24.7–33)

31.1 (27.7–35.8)

25.6(22.7–29)

< 0.001

35(29.6–41.7)

28(24.5–32.3)

< 0.001

Obesity: N(%)

1409 (39.55%)

1045 (57.5%)

364 (20.9%)

< 0.001

209 (72.6%)

1200 (36.6%)

< 0.001

Dyscholestrolemia: N(%)

1953 (54.8%)

1146 (63.0%)

807 (46.2%)

< 0.001

162 (56.2%)

1791 (54.7%)

0.609

Hypertension: N(%)

1044 (29.3%)

615 (33.8%)

429 (24.6%)

< 0.001

127 (44.1%)

917 (28.0%)

< 0.001

Diabetes: N(%)

889 (25.0%)

590 (32.5%)

299 (17.1%)

< 0.001

114 (39.6%)

775 (23.7%)

< 0.001

Median(quartile 1- quartile 3) was for continuous variables since they were non-normally distributed. P value was calculated by weighted logistic regression model. N (%) was for categorical variables. P value was calculated by weighted chi-square test.

Of all the 3563 participants, 1818 (51%) were diagnosed as NAFLD and 1745(49%) were non-NAFLD. Compared to non-NAFLD group, participants with NAFLD were inclined to be men, older, with higher CAP, LSM, TC, AST, ALT, GGT, uric acid, glycohemoglobin, ferritin, platelet count, and HS-CRP, but lower HDL-C and physical activity (p < 0.001). Participants with NAFLD had a higher level of BMI, waist circumference and higher propensity of obesity, hypertension, dyscholestrolemia, diabetes(P < 0.001). Especially, participants with NAFLD had a higher level of RBC count, hemoglobin, and RDW (P < 0.001).

Among the 3563 participants, 288(8%) had advanced liver fibrosis and 3275(92%) were without advanced liver fibrosis. Compared with those without advanced liver fibrosis, participants with advanced liver fibrosis were older, with higher proportion of men, with higher LSM, glycohemoglobin, AST, ALT, GGT, SC, ferritin and HS-CRP. Participants without advanced liver fibrosis had lower CAP, TC, serum albumin, platelet count and HDL-C than those with advanced liver fibrosis(p < 0.001). A higher level of BMI, waist circumference and higher propensity of obesity, hypertension, diabetes than those without advanced liver fibrosis were discovered in participants with advanced liver fibrosis(P < 0.001), but not dyscholestrolemia (P = 0.609). In addition, participants with advanced liver fibrosis had a higher level of RDW than those without advanced liver fibrosis (P < 0.001), but there was no significant difference were found in the two groups between the levels of RBC count, hemoglobin.

The Associations between RBC markers and NAFLD and CAP

As shown in Table 2, the associations of RBC count, MCHC and RDW with NAFLD and CAP were detected. After fully adjusting potential confounders, RBC count was positively associated with NAFLD (OR = 1.5, 95% CI = 1.2–1.8, P < 0.001) and CAP (β = 12.1, 95% CI = 8.1–16.2, p < 0.001).CAP values or the risk of NAFLD increased more progressively in the higher quartile groups of RBC count when compared to the lowest quartile group (P for trend < 0.001).After fully adjusting potential confounders, MCHC was also positively associated with NAFLD (OR = 1.2, 95% CI = 1.1–1.3, P = 0.003) and CAP (β = 2.7, 95% CI = 0.6–4.8, p = 0.013). CAP values or the risk of NAFLD significantly increased in the higher quartile groups of MCHC when compared to the lowest quartile group (the risk of NAFLD: P for trend < 0.001;CAP values: p for trend = 0.05).However, no significant difference was detected between RDW and the risk of NAFLD(OR = 1.0, 95% CI = 0.9–1.1, P = 0.688)or CAP(β = 0.8, 95% CI= -0.8 to 2.4, P = 0.307).

Table 2

The association of RBC markers with NAFLD and CAP(dB/m)

 

NAFLD

CAP(continuous)

 

Model 1

OR(95% CI) P value

Model 2

OR(95% CI) P value

Model 3

OR(95% CI) P value

Model 1

β(95% CI) P value

Model 2

β(95% CI) P value

Model 3

β(95% CI) P value

RDW

1.1 (1.0, 1.1) 0.011

1.1(1.0, 1.2) < 0.001

1.0 (0.9, 1.1) 0.688

2.0 (0.5, 3.5) 0.008

3.1 (1.6, 4.6) < 0.001

0.8 (-0.8, 2.4) 0.307

RBC count

1.7 (1.5, 1.9) < 0.001

1.9(1.6, 2.2) < 0.001

1.5 (1.2, 1.8) < 0.001

20.9(17, 24.9) < 0.001

22.4 (18.1, 26.7) < 0.001

12.1 (8.1, 16.2) < 0.001

Q1(2.8–4.4)

reference

reference

reference

reference

reference

reference

Q2(4.4–4.7)

1.3 (1.1, 1.6) 0.007

1.4 (1.1, 1.6) 0.002

1.3 (1.0, 1.6) 0.034

7.7 (2.0, 13.4) 0.008

8.3 (2.7, 13.9) 0.004

6.3 (1.3, 11.2) 0.013

Q3(4.7-5.0)

1.7 (1.4, 2.0) < 0.001

1.8 (1.4, 2.2) < 0.001

1.5 (1.2, 1.9) < 0.001

19.3 (13.7, 25.0) < 0.001

18.8 (13.1, 24.5) < 0.001

13.7 (8.6, 18.7) < 0.001

Q4 (5.0-7.8)

2.1 (1.7, 2.5) < 0.001

2.3 (1.8, 2.8) < 0.001

1.6 (1.2, 2.0) < 0.001

27.6 (21.9, 33.2) < 0.001

28.0 (21.9, 34.2) < 0.001

15.7 (10.3, 21.2) < 0.001

P for trend

< 0.001

< 0.001

< 0.001

< 0.001

< 0.001

< 0.001

MCHC

1.1 (1.0, 1.2) 0.002

1.0 (1.0, 1.1) 0.379

1.2 (1.1, 1.3) 0.003

3.7 (1.5, 5.8) < 0.001

0.4 (-1.8, 2.6) 0.719

2.7 (0.6, 4.8) 0.013

Q1 (25.2–32.9)

reference

reference

reference

reference

reference

reference

Q2 (33.0-33.5)

1.1 (0.9, 1.3) 0.534

0.9 (0.8, 1.1) 0.549

1.1 (0.9, 1.4) 0.475

1.6 (-4.2, 7.4) 0.594

-3.2 (-9.0, 2.5) 0.269

1.3 (-3.7, 6.3) 0.601

Q3 (33.6–34.0)

1.1 (0.9, 1.4) 0.245

1.0 (0.8, 1.2) 0.857

1.1 (0.9, 1.4) 0.264

3.5 (-2.4, 9.4) 0.243

-2.2 (-8.0, 3.7) 0.472

3.3 (-1.9, 8.4) 0.214

Q4 (34.1–38.3)

1.4 (1.1, 1.6) < 0.001

1.2 (1.0, 1.4) 0.130

1.5 (1.2, 1.9) < 0.001

9.3 (3.6, 15.0) 0.001

2.0 (-3.8, 7.8) 0.494

9.0 (4.0, 14.1) < 0.001

P for trend

< 0.001

0.094

< 0.001

< 0.001

0.377

0.050

Weighted logistic regression was used to detect the associations of NAFLD and CAP values withRBC count, MCV, MCHC and RDW. Model 1: unadjusted; Model 2: age, sex, race/ethnicity. Model 3: age, sex, race/ethnicity, smoking, obesity, diabetes, dyscholestrolemia, hypertension, physical activity, anemia.

The Associations between RBC markers and advanced liver fibrosis and LSM

As shown in Table 3, the associations of RBC count, MCHC and RDW with advanced liver fibrosis and LSM were detected. After fully adjusting potential confounders, RDW was positively associated with advanced liver fibrosis(OR = 1.3, 95% CI = 1.1–1.4, P < 0.001) and LSM(β = 0.2, 95% CI = 0.1–0.3, p = 0.005).LSM or the risk of advanced liver fibrosis increased more progressively in the higher quartile groups of RDW when compared to the lowest quartile group (the risk of advanced liver fibrosis: P for trend < 0.001; LSM: P for trend = 0.002 ).However, no significant difference was found between RBC count or MCHC with the risk of advanced liver fibrosis or LSM.

Table 3

The association of RBC markers with advanced liver fibrosis and LSM (kPa)

 

Advanced liver fibrosis

LSM(continuous)

 

Model 1

OR(95% CI) P value

Model 2

OR(95% CI) P value

Model 3

OR(95% CI) P value

Model 1

β(95% CI) P value

Model 2

β(95% CI) P value

Model 3

β(95% CI) P value

RBC count

1.1 (0.8, 1.4) 0.596

1.1 (0.8, 1.4) 0.648

0.8 (0.6, 1.1) 0.225

0.3 (-0.0, 0.6) 0.099

0.1 (-0.2, 0.5) 0.426

-0.1 (-0.5, 0.2) 0.543

MCHC

1.0 (0.9, 1.2) 0.844

1.0 (0.8, 1.1) 0.518

1.0 (0.9, 1.2) 0.930

-0.0 (-0.2, 0.1) 0.576

-0.1 (-0.3, 0.0) 0.110

-0.1 (-0.3, 0.1) 0.458

RDW

1.2 (1.1, 1.3) < 0.001

1.2 (1.1, 1.3) < 0.001

1.3 (1.1, 1.4) < 0.001

0.2 (0.1, 0.3) < 0.001

0.2 (0.1, 0.3) < 0.001

0.2 (0.1, 0.3) 0.005

Q1 (11.3–13.0)

reference

reference

reference

reference

reference

reference

Q2 (13.1–13.5)

1.5 (1.0, 2.3) 0.069

1.4 (0.9, 2.1) 0.146

1.2 (0.8, 1.9) 0.414

0.2 (-0.2, 0.7) 0.267

0.1 (-0.4, 0.5) 0.703

-0.0 (-0.5, 0.4) 0.837

Q3 (13.6–14.2)

1.8 (1.2, 2.7) 0.005

1.6 (1.1, 2.4) 0.026

1.3 (0.8, 2.0) 0.249

0.4 (0.0, 0.9) 0.050

0.3 (-0.2, 0.7) 0.257

0.0 (-0.4, 0.5) 0.895

Q4 (14.3–29.2)

3.0 (2.1, 4.5) < 0.001

2.8 (1.9, 4.2) < 0.001

2.3 (1.5, 3.5) < 0.001

1.3 (0.8, 1.7) < 0.001

1.2 (0.7, 1.6) < 0.001

0.8 (0.3, 1.2) < 0.001

P for trend

< 0.001

< 0.001

< 0.001

< 0.001

< 0.001

0.002

Weighted logistic regression was used to detect the associations of advanced liver fibrosis and LSM with RBC count, MCV, MCHC and RDW. Model 1: unadjusted; Model 2: age, sex, race/ethnicity. Model 3: age, sex, race/ethnicity, smoking, obesity, diabetes, dyscholestrolemia, hypertension, physical activity, anemia.

Discussion

The present study elicited several findings. We revealed a significantly positive association between RBC count, MCHC and the risk of NAFLD. On the whole, every one-unit (1012/L) increase in RBC count was associated with a 50% increased risk for NAFLD while everyone-unit (g/dL) increase in MCHC was related with a 20% increased risk for NAFLD. Participants in higher quartiles of RBC counts and MCHC had a higher risk for NAFLD. Moreover, RDW was positively associated with advanced liver fibrosis and every one-unit (%) increase in RDW was related with a 30% increased risk for NAFLD. Participants in higher quartiles of RDW had a higher risk for advanced liver fibrosis. However, no correlation were detected between RBC counts and advanced liver fibrosis, MCHC and advanced liver fibrosis, RDW and NAFLD.

A limited number of studies have focused on the association between RBC and NAFLD, and the results remain controversial and inconclusive. Wang H et al. conducted a large-scale cross-sectional study of 8618 southern Chinese adults, discovering higher RBC count was positively associated with a higher risk of ultrasonography-diagnosed fatty liver disease[10], but their results were puzzled by alcohol consumption. A cohort study that included 27,112 subjects with up to 5 years of follow-up in China indicated increased RBC count might precede the onset of NAFLD[11]. The same results were found in another study[19]. However, Danny Issa et al. concluded no significant association was detected between the RBC count and severity degree of NAFLD depended on the histopathological features [20]. This inconsistency among these studies might be due to the heterogeneity among sample sizes, study designs, samples from different regions, and methods to detect NAFLD. We innovatively found the intimate relationship of MCHC and NAFLD, since no literature has previously reported. In the present study, we revealed RBC count and MCHC were positively related with the risk of NAFLD but not liver fibrosis in the general American population.

In the present study, we revealed RDW was positively associated with advanced liver fibrosis, but not associated with NAFLD. Several previous studies investigating the correlation of RDW with NAFLD and advance fibrosis have been also published over the last decade [12, 13, 21]. A retrospective study including 24,547 subjects from Korea analyzed the association between RDW values and the degree of fibrosis in NAFLD[13]. NAFLD was diagnosed on abdominal ultrasonography and the degree of liver fibrosis was determined according to BARD and FIB-4 scores. After adjusting for age, hemoglobin level, mean corpuscular volume, history of hypertension, history of diabetes, and high-sensitivity C-reactive protein, elevated RDW was found independently associated with advanced fibrosis in NAFLD. Although the ethnicity, sample size, diagnostic approach, and adjusted confounding factors are different, the result was congruent with that in our study. Yang et al collected data from 1637 normal control individuals and 619 NAFLD patients for routine medical check-up in their hospital and found that patients with NAFLD defined by abdominal ultrasonography were more likely to have high levels of RDW (β = 0.301, P < 0.01) analyzed by binary linear regression analysis. The independent variable was NAFLD and the dependent variable was the level of RDW in their study. The inconsistency in this study with our result might mainly caused by the differences in ethnicities of participants, sample size and statistical approach. More detailed study should be carried on to determine the correlation of RDW with NAFLD and advance fibrosis in general population of Americans.

Previous literature has preliminarily explained how RBCs function in the pathogenesis of NAFLD and progression. RBCs are well equipped with antioxidant systems, which can protect liver tissue from injury and regulate cardiovascular homeostasis while NAFLD is frequently accompanied by iron overload, inflammation, and oxidative damage[22].Oxidative stress in metabolic syndrome and NAFLD could enhance compensatory RBC count, functioning to remedy liver function. Iron is proved vital to the creation of hemoglobin and red blood cell, therefore iron overload and higher ferritin levels in NAFLD might also lead to the increase of compensatory RBC count in circulation[23]. Besides, the capacity of the liver to degrade aging RBCs in NAFLD patients might be lower. Besides, the liver is an important organ for metabolism and regularly produces various pro-oxidant reactive species. When RBCs flow through liver tissues, the liver functions to eliminate aging RBCs, resulting in the release of iron and triggering oxidative stress[10]. However, further thorough studies focusing on the mechanism for higher RBC count in NAFLD are urgently needed with fatty liver biopsy samples from patients or in animal and cell models.

Due to the nationally representative nature of NHANES, the strength of our findings lies in the large size of the study samples. Focusing on the correlation of RBC count and the risk of NAFLD, this study included 3563 participants, the largest number of study samples in our perspective. However, there were some shortcomings. First, NAFLD was confirmed based on CAP detected by FibroScan® but not the gold standard liver biopsy and many participants lacking available VCTE data were excluded, which might cause bias in including NAFLD participants. Moreover, the accuracy needs further checked with CAP ≥ 285dB/m to diagnose NAFLD. Second, patients with known hepatitis B and C as well as those with significant alcohol intake were ruled out. Nevertheless, several other liver diseases might have been underestimated and not been considered (e.g. AIH, cholestatic diseases). Third, in a large sample size, even minor and non-significant differences come as significant p value. Forth, though we have adjusted several important covariates, other potential factors might introduce the bias to induce the conclusion not credible.

Conclusion

We revealed that higher RBC count and MCHC were associated with NAFLD while elevated RDW was associated with advanced liver fibrosis in Americans. These findings facilitate identifying RBC count and MCHC as convenient and effective predictors for risk assessment for NAFLD and RDW as a potential index in algorithms for liver fibrosis risk prediction, respectively.

Abbreviations

NAFLD

Nonalcoholic fatty liver disease

MetS

metabolic syndrome

RBC

red blood cell

MCHC

corpuscular hemoglobin concentration

RDW

red blood cell distribution width

NHANES

National Health and Nutrition Examination Survey

CAP

controlled attenuation parameter

LSM

liver stiffness measurement

MEC

mobile examination center

VCTE

vibration control transient elastography

BMI

body mass index

PIR

poverty income ratio

TC

total cholesterol

LDL-C

low-density lipoprotein cholesterol

HDL-C

high-density lipoprotein cholesterol.

Declarations

Acknowledgements
 
The authors thank the staff and the participants of the NHANES study for their valuable contributions.

Author contributions

YWC conceived this study design. TG and NBY were in charge of acquisition and analysis of data. TG and NBY drafted, revised, and critically reviewed the article. All authors approved the version to be published on the journal, and accountable for all aspects of the work.

Funding

This work was financially supported by grants from the National Natural Science Foundation of China (81970511); the First Batch of Young Technical Backbone Talents Project of Ningbo Municipal Health Commission; and TianQing Liver Diseases Research Fund Subject of Chinese Foundation for Hepatitis Prevention and Control (TQGB20180358). 

Availability of data and materials
 The data are publicly available on the Internet (http://www.cdc.gov/nchs/nhanes/). 
Ethics approval and consent to participate

The ethics review board of the National Center for Health Statistics approved all NHANES protocols and written informed consent was obtained from all participants.

Consent for publication
 Not applicable.
Competing interests
 There is no conflict of interest between authors.

References

  1. Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M. Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology 2016; 64:73–84.
  2. Friedman SL, Neuschwander-Tetri BA, Rinella M, Sanyal AJ. Mechanisms of NAFLD development and therapeutic strategies. Nat Med 2018; 24:908–922.
  3. Kanwal F, Shubrook J, Adams L, et al. Clinical Care Pathway for the Risk Stratification and Management of Patients With Nonalcoholic Fatty Liver Disease. Gastroenterology 2021; 161:1657–1669.
  4. Powell EE, Wong VW, Rinella M. Non-alcoholic fatty liver disease. Lancet (London, England) 2021; 397:2212–2224.
  5. Tilg H, Moschen A. Evolution of inflammation in nonalcoholic fatty liver disease: the multiple parallel hits hypothesis. Hepatology (Baltimore, Md) 2010; 52:1836–1846.
  6. Kuhn V, Diederich L, Keller T, et al. Red Blood Cell Function and Dysfunction: Redox Regulation, Nitric Oxide Metabolism, Anemia. Antioxidants & redox signaling 2017; 26:718–742.
  7. Crawford J, Chacko B, Kevil C, Patel R. The red blood cell and vascular function in health and disease. Antioxidants & redox signaling 2004; 6:992–999.
  8. Huang L, Dou D, Liu N, et al. Association of erythrocyte parameters with metabolic syndrome in the Pearl River Delta region of China: a cross sectional study. BMJ open 2018; 8:e019792.
  9. Wu S, Lin H, Zhang C, et al. Association between erythrocyte parameters and metabolic syndrome in urban Han Chinese: a longitudinal cohort study. BMC public health 2013; 13:989.
  10. Wang H, Zhang H, Wu S, et al. Red blood cell count has an independent contribution to the prediction of ultrasonography-diagnosed fatty liver disease. PloS one 2017; 12:e0172027.
  11. Zhong F, Guan L, Lin H, et al. Red Blood Cell Count: An Unrecognized Risk Factor for Nonalcoholic Fatty Liver Disease. Frontiers in endocrinology 2021; 12:760981.
  12. Yang W, Huang H, Wang Y, Yu X, Yang Z. High red blood cell distribution width is closely associated with nonalcoholic fatty liver disease. European journal of gastroenterology & hepatology 2014; 26:174–178.
  13. Kim HM, Kim BS, Cho YK, et al. Elevated red cell distribution width is associated with advanced fibrosis in NAFLD. Clinical and molecular hepatology 2013; 19:258–265.
  14. Yao X, Xu X, Jin F, Zhu Z. The Correlation of Type 2 Diabetes Status with Bone Mineral Density in Middle-Aged Adults. Diabetes, metabolic syndrome and obesity : targets and therapy 2020; 13:3269–3276.
  15. Eddowes P, Sasso M, Allison M, et al. Accuracy of FibroScan Controlled Attenuation Parameter and Liver Stiffness Measurement in Assessing Steatosis and Fibrosis in Patients With Nonalcoholic Fatty Liver Disease. Gastroenterology 2019; 156:1717–1730.
  16. Vilar-Gomez E, Nephew L, Vuppalanchi R, et al. High-quality diet, physical activity, and college education are associated with low risk of NAFLD among the US population. Hepatology (Baltimore, Md) 2021.
  17. Statistics. NCfH. National Health and Nurtrition Examination Survey 2017–2018 Laboratory Data. Internet: https://wwwcdcgov/nchs/nhanes/Search/DataPageaspx?Component=Laboratory&CycleBegin (accessed June 2020) 2017.
  18. ADA. Classification and diagnosis of diabetes: standards of medical care in diabetes-2020. Diabetes care 2020; 43(Supplement 1)::S14-S31.
  19. Jiang Y, Zeng J, Chen B. Hemoglobin combined with triglyceride and ferritin in predicting non-alcoholic fatty liver. Journal of gastroenterology and hepatology 2014; 29:1508–1514.
  20. Issa D, Albeldawi M, Lopez R, Alkhouri N. Hematocrit levels and red blood cell indices in patients with nonalcoholic fatty liver disease. European journal of gastroenterology & hepatology 2014; 26:363–364.
  21. Yuyun D, Zhihua T, Haijun W, et al. Predictive value of the red blood cell distribution width-to-platelet ratio for hepatic fibrosis. Scandinavian journal of gastroenterology 2019; 54:81–86.
  22. Pawlak M, Lefebvre P, Staels B. Molecular mechanism of PPARα action and its impact on lipid metabolism, inflammation and fibrosis in non-alcoholic fatty liver disease. Journal of hepatology 2015; 62:720–733.
  23. Yang N, Lu Y, Cao L, Lu M. The association between non-alcoholic fatty liver disease and serum ferritin levels in American adults. Journal of Clinical Laboratory Analysis; n/a:e24225.