Patients with Nonalcoholic Steatohepatitis and Advanced Liver Disease Have the Lowest Cardiorespiratory Fitness

Abstract

Background & Aims

Cardiorespiratory fitness and liver fibrosis are independently associated with poor outcomes in patients with nonalcoholic steatohepatitis (NASH), however, conflicting reports exist about their relationship. We aimed to better characterize the relationship between cardiorespiratory fitness and liver histology in a cross-sectional study of patients with biopsy-proven NASH.

Methods

Participants aged 18–75 years completed VO_2peak fitness assessment using symptom-limited graded exercise testing. Participants were compared by liver fibrosis stage and NAFLD Activity Score (NAS). Multivariable models were constructed to assess factors related to relative VO_2peak, including liver fibrosis and NAS.

Results

Thirty-five participants with mean age 48 ± 12 years and body mass index 33.5 ± 7.6 kg/m² were enrolled. Seventy-four percent of participants were female and 49% had diabetes. A dose-dependent relationship was found between relative VO_2peak and liver fibrosis. Relative VO_2peak was significantly lower in participants with advanced fibrosis (F3 disease-15.7 ± 5.3 vs. ≤ F2 disease- 20.7 ± 5.9 mL/kg/min, p = 0.027). NAS > 5 was also associated with lower relative VO_2peak (22.6 ± 5.7 vs. 16.5 ± 5.1 mL/kg/min, p = 0.012) compared to NAS ≤ 5. With multivariable modeling, advanced fibrosis remained independently predictive of relative VO_2peak while NAS trended towards significance.

Discussion and Conclusions

Advanced liver fibrosis is independently associated with cardiorespiratory fitness in patients with NASH. This may explain the incremental increase in mortality as liver fibrosis stage increases. Further research is needed to determine if exercise training can improve cardiorespiratory fitness across multiple stages of liver fibrosis and directly reduce morbidity and mortality in patients with NASH.

Introduction

Nonalcoholic fatty liver disease (NAFLD) is a leading cause of chronic liver disease worldwide affecting nearly 30% of the global population [1]. NAFLD, and its more severe variant nonalcoholic steatohepatitis (NASH), develop in part due to physical inactivity [2]. Compared to the general population, patients with NAFLD have lower survival, independent of comorbid metabolic risk factors [3]. There is an incremental increase in mortality with each liver fibrosis stage (F) in NASH [3, 4]. Patients with NASH and advanced fibrosis (F3) have nearly two times greater overall and liver-related mortality when compared to patients with F2 disease [4]. Additionally, NASH patients with advanced fibrosis have increased death from cardiovascular disease (CVD) and hepatocellular carcinoma (HCC) [5–7]. Why patients with F3 disease develop greater rates of CVD and HCC remains a highly significant knowledge gap.

Maximal oxygen uptake (VO_2peak) is a clinical assessment of cardiorespiratory fitness that measures oxygen consumption during incremental, graded exercise testing using treadmill, cycle ergometer or arm ergometer [8]. VO_2peak varies across age and sex with accepted normative ranges and fitness level categorizations (Table 1) [8]. Patients with NASH have lower cardiorespiratory fitness than the general population [9, 10]. Most patients with NASH have a poor or very poor fitness level, independent of age, body mass index (BMI), and sex [11]. Cardiorespiratory fitness is predictive of mortality in the general population and in those with chronic disease, including NAFLD [12–15]. The HUNT study found all-cause mortality was increased 52% in NAFLD patients with low cardiorespiratory fitness compared to participants with high cardiorespiratory fitness [15].

Table 1.

Sex-specific measured VO_2peak from treadmill and cycle exercise tests obtained from FRIEND

Treadmill testing	Age group (years) Fitness level	Percentile
		5th	10th	25th	50th	75th	90th	95th
		Very poor		Poor	Average	Good	Excellent
Men	20–29	29.0	32.1	40.1	48.0	55.2	61.8	66.3
	30–39	27.2	30.2	35.9	42.4	49.2	56.5	59.8
	40–49	24.2	26.8	31.9	37.8	45.0	52.1	55.6
	50–59	20.9	22.8	27.1	32.6	39.7	45.6	50.7
	60–69	17.4	19.8	23.7	28.2	34.5	40.3	43.0
	70–79	16.3	17.1	20.4	24.4	30.4	36.6	39.7
Women	20–29	21.7	23.9	30.5	37.6	44.7	51.3	56.0
	30–39	19.0	20.9	25.3	30.2	36.1	41.4	45.8
	40–49	17.0	18.8	22.1	26.7	32.4	38.4	41.7
	50–59	16.0	17.3	19.9	23.4	27.6	32.0	35.9
	60–69	13.4	14.6	17.2	20.0	23.8	27.0	29.4
	70–79	13.1	13.6	15.6	18.3	20.8	23.1	24.1

Cycle ergometer testing	Age group (years) Fitness level	Percentile
		10th	20th	50th	70th	90th
		Very Poor	Poor	Average	Good	Excellent
Men	20–29	29.5	33.2	41.9	47.9	55.5
	30–39	21.8	25.4	30.1	33.9	41.7
	40–49	20.6	22.2	28.6	30.4	47.1
	50–59	20.4	21.5	26.3	28.2	34.0
	60–69	17.3	19.0	23.2	24.5	29.9
	70–79	15.8	16.7	20.4	21.9	28.1
Women	20–29	19.3	21.6	31.0	35.6	42.6
	30–39	20.9	17.0	21.6	24.2	30.0
	40–49	14.6	15.8	19.4	22.0	26.2
	50–59	13.7	14.9	17.3	19.3	22.6
	60–69	13.0	14.0	16.0	17.8	20.5
	70–79	12.0	12.8	14.8	16.9	18.0

Adapted from Kaminsky et al. [8] FRIEND Fitness Registry and the Importance of Exercise National Database

The relationship between fibrosis stage and cardiorespiratory fitness in NASH has been explored by several previous studies [10, 16]. Krasnoff et al. performed a cross-sectional analysis of 37 patients with NAFLD, 81% (n = 30) of whom had NASH [10]. While the authors demonstrated VO_2peak was significantly lower for NASH participants, including those with greater histologic activity defined by NAFLD Activity Score (NAS) > 4, they did not find any measurable difference in their subgroup analysis comparing participants with F1 (n = 25) to participants with F2/3 (n = 12). Canada et al. measured cardiorespiratory fitness in 36 patients with NAFLD and found VO_2peak was inversely related to liver fibrosis stage, however, 42% of included participants had NAFL and no subgroup analysis was performed for only those participants with NASH, limiting the authors’ conclusions about the relationship between liver fibrosis stage and cardiorespiratory fitness [16].

Given the conflicting reports about the relationship between liver fibrosis stage and cardiorespiratory fitness as well as the significant heterogeneity in the study populations examined, this is a highly significant area of study with key unanswered questions. We aimed to better characterize the relationship between liver fibrosis stage and cardiorespiratory fitness in patients exclusively with biopsy-proven NASH. We hypothesized that patients with advanced liver fibrosis have lower cardiorespiratory fitness independent of other predictors of poor fitness including age, BMI and sex.

Materials and Methods

Study Cohort

This is a post hoc analysis of two clinical trials [17, 18] in which sedentary adults completing < 90 min/week of moderate-intensity exercise [equivalent of 3–6 metabolic equivalents (MET), per self-report over the previous three months] with biopsy-proven NASH were recruited and enrolled at two US-based tertiary care academic centers and underwent exercise testing as a part of each study protocol. NASH was defined as ≥ 5% steatosis plus hepatocyte ballooning and/or lobular inflammation [19]. NAS, a summary score including individual components of steatosis (0–3), lobular inflammation (0–3) and hepatocyte ballooning (0–2), was calculated for each participant by a blinded liver pathologist and reviewed by a blinded hepatologist [20]. Based on previous studies, patients were included with NAS > 4. Liver histology collected within the six months preceding enrollment was required [10, 21, 22].

Patients with cirrhosis were excluded. Patients were also excluded if other causes of liver disease were identified by standard serologic evaluation or if alcohol consumption was greater than 21 drinks per week for men or 14 drinks per week for women. Additionally, patients were excluded for active smoking, BMI < 18 kg/m² or > 45 kg/m², secondary causes of hepatic steatosis (e.g., lipodystrophy, long-term use of steatogenic medications, monogenic hereditary disorders, parenteral nutrition), severe psychiatric illness, advanced medical comorbidities, or pregnancy. Any patient with contraindications to safely undergoing graded exercise testing was also excluded (e.g., abnormal electrocardiogram, active cardiopulmonary symptoms, restricted range of motion or mobility). Patients were excluded if they had participated in an interventional NASH clinical trial in the preceding six months. Both the Pennsylvania State University College of Medicine Human Subjects Protection Office/Institutional Review Board (IRB) and the University of Virginia IRB for Health Sciences Research approved all parts of this study protocol. Informed consent was obtained for each participant by trained study staff.

Demographic and Laboratory Data

Participant interviews and review of the electronic medical record by study staff were completed to obtain all relevant demographic information. Anthropometric, biochemical, and clinical assessments were performed. Laboratory testing including alanine aminotransferase (ALT), aspartate aminotransferase (AST), fasting glucose, and fasting lipid measurements [e.g., high-density lipoprotein (HDL), low-density lipoprotein (LDL), triglyceride and total cholesterol] were performed.

Cardiopulmonary Fitness Testing

Participants underwent VO_2peak fitness testing using a symptom-limited continuous incremental exercise protocol via cycle ergometer or treadmill based on study site experience, preference, and pre-testing assessment. Prior to exercise testing, participants were instructed to avoid strenuous activity and hold medications with AV nodal action including beta blockers for 24 h. Participants were also instructed to not intake caffeine or alcohol for at least the preceding 12 h and were instructed to be fasting for at least four hours. Exercise testing, utilizing a standardized protocol that incrementally adjusted the speed, incline, or resistance, was performed in the presence of both a physician and an American College of Sports Medicine (ACSM)-certified exercise physiologist. Participants were instructed to pedal or run until they were physically unable, deemed unsafe to continue exercising, or could not maintain the required treadmill speed or pedal cadence. The Bruce protocol was utilized for treadmill testing [23]. For the cycle ergometer testing, the participant started at a power output of 20 watts. Every three minutes the power output was increased by 15 watts until exhaustion was reached. For treadmill testing, the participant was instructed on how to use the Borg Rated of Perceived Exertion (RPE) scale (range 6–20) [24]. Based on our previous experience, VO_2peak was determined by the highest oxygen uptake with maximal heart rate (HR) within 10 beats per minute (bpm) of age-predicted HR max (220 bpm—age in years), a respiratory exchange ratio (RER) value > 1.05, and/or an RPE > 18 [11]. VO_2peak was determined by data averaged across 30 s. ParvoMedics TrueOne2400 metabolic measuring system was utilized for indirect calorimetry and VO_2peak measurement. Continuous electrocardiogram monitoring was performed and blood pressure was measured at every stage of the protocol, unless the participant began to run for the graded treadmill test. Where appropriate, verbal encouragement was provided by study staff administering the cardiorespiratory fitness test. After completion of VO_2peak testing, a standard cool down period ensued for up to five minutes. For all maximal tests, discontinuation was symptom limited. Relative VO_2peak was expressed in mL/kg/min.

Statistical Analysis

Histologic fibrosis stage F3 was chosen as the categorical cut-off value based on previous studies correlating advanced fibrosis with CVD, HCC and death [4–7]. Accordingly, participants with advanced fibrosis (F3) were compared to participants without advanced fibrosis (≤ F2). Standard univariate analysis was performed for categorical and continuous variables as appropriate using Student’s t test, Mann–Whitney U test, chi-square test or Fisher-exact test. Multivariable models were constructed using linear regression to assess factors related to relative VO_2peak. Covariates that were deemed clinically significant or statistically significant with p value < 0.10 were included in the final model. Final variables included in the model were fibrosis stage, age in years, BMI, diabetes, female sex and NAS (steatosis was not included in the final model despite its statistical significance as it is a component of NAS). A p value of 0.05 was considered significant. All statistical analysis was completed using SAS Version 9.4 (Cary, NC). Graphs were constructed with GraphPad Prism Version 7.0 (San Diego, CA).

Results

Overall Study Cohort

Thirty-five patients (74% female) with mean age 48 ± 12 years were enrolled. Forty-nine percent of participants were diabetic, 46% had hypertension and 46% had hyperlipidemia. Mean BMI was 33.5 ± 7.6 kg/m² and mean NAS was 5.7 ± 0.8. Mean liver fibrosis stage was 1.9 ± 0.9 with individual stages as follows: F0 (n = 2), F1 (n = 11), F2 (n = 12) and F3 (n = 10). Individual NAS components were steatosis grade 0/1/2/3 (n = 0/4/14/17), lobular inflammation 0/1/2/3 (n = 1/12/12/10) and hepatocyte ballooning 0/1/2 (n = 2/15/18). Summative NAS scores were 5/6/7/8 (n = 0/16/12/7). Twenty-one participants completed graded treadmill testing and 14 underwent cycle ergometer testing. RER was similar between the two groups of testing indicating that either modality was successful in achieving a maximal cardiopulmonary fitness assessment based on the a priori criteria (Supplementary Tables 1 & 2). Mean relative VO_2peak was 19.0 ± 6.0 mL/kg/min for the entire cohort. Almost all participants had “poor” (14%) or “very poor” (71%) fitness level based on age and sex-adjusted ranges (Table 1) as appropriate for testing modality.

Comparison of Fibrosis Stage

Using the a priori determined F3 histologic cut-off [5–7], participants with advanced fibrosis (n = 10) were compared to those without advanced fibrosis (n = 25) and no clinically or statistically important differences were observed in baseline characteristics, including established predictors of cardiorespiratory fitness such as age, BMI and sex (Table 2). Relative VO_2peak was significantly lower in the advanced fibrosis group (15.7 ± 5.3 vs. 20.7 ± 5.9 mL/kg/min, p = 0.030) (Fig. 1). When comparing participants across all liver fibrosis stages, a dose-dependent relationship was also observed with F0/1 individuals having the greatest cardiorespiratory fitness, followed by F2 then F3 (Fig. 2). Supplementary Table S3 shows cardiorespiratory fitness testing characteristics by liver fibrosis stage.

Table 2.

Baseline participant characteristics based on fibrosis stage cohort

	No advanced liver fibrosis ≤ F2 (n = 25)	Advanced liver fibrosis F3 (n = 10)	p value
Age (year), mean (95% CI)	46 (41–52)	53 (46–60)	0.144
Female, n (%)	18 (72)	8 (80)	0.625
Body weight (kg), mean (95% CI)	100 (93–107)	95 (85–106)	0.450
Body mass index (kg/m2), mean (95% CI)	34.1 (32.3–35.9)	35.5 (29.8–40.7)	0.576
Waist-circumference (in), mean (95% CI)	45.1 (43.3–46.9)	44.1 (39.6–48.6)	0.608
Diabetes, n (%)	13 (52)	4 (40)	0.521
Hypertension, n (%)	11 (44)	5 (50)	0.748
ALT (IU/L), mean (95% CI)	64 (51–77)	69 (44–95)	0.694
AST (IU/L), mean (95% CI)	53 (35–70)	66 (41–91)	0.361
Fasting glucose (mg/dL), mean (95% CI)	126 (109–144)	122 (104–140)	0.755
HDL cholesterol (mg/dL), mean (95% CI)	44 (40–48)	41 (35–46)	0.363
LDL cholesterol (mg/dL), mean (95% CI)	115 (101–129)	132 (115–150)	0.140
Total cholesterol (mg/dL), mean (95% CI)	197 (181–213)	204 (187–221)	0.624
Triglycerides (mg/dL), mean (95% CI)	239 (152–325)	176 (123–228)	0.362
NAS, mean (95% CI)	5.7 (5.2–6.2)	5.8 (5.4–6.1)	0.841
Steatosis	2.6 (2.3–2.8)	2.0 (1.4–2.6)	0.029
Ballooning	1.6 (1.4–2.1)	2.2 (1.6–2.8)	0.179
Inflammation	1.4 (1.2–1.8)	1.6 (1.1–1.8)	0.797

ALT alanine aminotransferase, AST aspartate aminotransferase, CI confidence interval, HDL high-density lipoprotein, LDL low-density lipoprotein, NASH nonalcoholic steatohepatitis, NAS NASH activity score In general, the two fibrosis cohorts were similar with no clinically or statistically significant differences

Fig. 1.

Patients with NASH and advanced liver fibrosis have the poorest cardiorespiratory fitness. Independent of traditional risk factors for poor fitness, participants with advanced fibrosis had lower relative VO_2peak compared to those without advanced fibrosis (15.7 ± 5.3 vs. 20.7 ± 5.9 mL/kg/min)

Fig. 2.

Comparison of cardiorespiratory fitness amongst all fibrosis stages in patients with NASH. A dose-dependent relationship was observed between relative VO_2peak and liver fibrosis stage where participants with F0/F1 had the greatest cardiorespiratory fitness, followed by F2 then F3

Comparison of NAS

Given the relationship between inflammation and physical activity, participants were compared across different NAS scores. Those with NAS > 5 were significantly more likely to have lower relative VO_2peak when compared to participants with NAS ≤ 5 (22.6 ± 5.7 vs. 16.5 ± 5.1 mL/kg/min, p = 0.012) (Fig. 3).

Fig. 3.

Histologic NASH activity is associated with cardiorespiratory fitness in patients with NASH. Participants with NAS > 5 were significantly more likely to have lower relative VO_2peak when compared to those with NAS ≤ 5 (22.6 ± 5.7 vs. 16.5 ± 5.1 mL/kg/min)

Predictors of Cardiopulmonary Fitness

Using multivariable modeling, only liver fibrosis stage remained as an independent predictor of cardiorespiratory fitness in participants with biopsy-proven NASH. Adjusted multivariable modeling found advanced fibrosis was independently predictive of relative VO_2peak (β − 4.04, 95% CI − 7.71 to − 0.37, p = 0.027) where individuals with F3 were over four-times more likely to have a 5% reduction in aerobic capacity when compared to those with F ≤ 2 (Table 3). NAS approached, but did not reach, statistical significance.

Table 3.

Predictors of VO_2peak in participants with biopsy-proven NASH

	β-coefficient	95% Confidence interval	p value
Age (years)	− 0.12	− 0.27 to 0.04	0.126
BMI (kg/m2)	− 0.18	− 0.49 to 0.12	0.232
Diabetes	− 1.65	− 4.97 to 1.78	0.340
Female	− 3.61	− 7.92 to 0.69	0.097
Advanced fibrosis (F3)	− 4.04	− 7.71 to − 0.37	0.032
NAFLD activity score	− 2.22	− 4.49 to 0.05	0.055

Multivariable linear regression modeling demonstrates advanced liver fibrosis (F3) is predictive of lower cardiorespiratory fitness level in participants with NASH independent of traditional risk factors

Discussion

This is the first study to investigate the relationship between fibrosis stage and cardiorespiratory fitness in a well-phenotyped cohort exclusively consisting of patients with biopsy-proven NASH. Previous studies included NAFLD patients across all stages of disease, many of whom did not have NASH [10, 16]. Our findings extend our understanding of the relationship between cardiorespiratory fitness and liver fibrosis stage in patients with NASH in three ways. First, there was an independent relationship between advanced fibrosis and V^·O_2peak when adjusting for common determinants of cardiorespiratory fitness including age, sex, obesity and diabetes. Second, cardiorespiratory fitness levels may be even lower than previously thought for patients with NASH; 85% of the study population had either poor or very poor age- and sex-adjusted VO_2peak. Third, we demonstrated that NASH histologic activity is also related to cardiorespiratory fitness where patients with NAS > 5 had the lowest VO_2peak.

Our finding that cardiorespiratory fitness is lowest in participants with NASH and advanced liver fibrosis challenges the way we treat patients with NASH. Currently, each patient is approached similarly regardless of the stage of liver fibrosis and both lifestyle change, including exercise training, and pharmacologic therapy are recommended. Our findings raise the question of whether patients with NASH and advanced liver fibrosis will respond to exercise training or are they too deconditioned to gain enough benefit from exercise training to not only significantly improve their poor cardiorespiratory fitness but lessen their fibrosis stage? Focusing efforts on implementing exercise training regimens in patients with earlier-stage disease (≤ F2) may be considered with these findings, allowing pharmacologic therapy, including experimental medications in clinical trials, to be directed towards patients with the most advanced liver fibrosis. Alternatively, since patients with advanced liver fibrosis are more affected in their cardiorespiratory fitness, they could show the greatest capacity for a relative improvement making them an ideal target for appropriately supported exercise training. Comparative studies considering healthcare utilization would be needed to resolve such dilemma.

The findings that patients with NASH have lower VO_2peak than previously thought is important because low cardiorespiratory fitness is associated with increased mortality [15]. While further study is needed to better understand this relationship, it could be a key link in explaining the incremental increase in mortality as liver fibrosis stage increases in NASH [3, 4]. Poor fitness level and low physical activity are well-described risk factors for both CVD and cancer (including HCC), the two leading causes of death for patients with NAFLD and NASH. Additionally, the greatest risk of CVD and HCC in NASH patients without cirrhosis is seen for those with F3 disease [5–7]. Previous studies in non-NASH patients have shown that small gains in cardiopulmonary fitness can lead to improvements in coronary artery disease parameters and mortality [25–27]. For example, Imboden et al. appreciated a 12% decline in all-cause mortality with each interval increase in metabolic equivalent in a sample of reportedly healthy participants [27].

To our knowledge, the direct impact of exercise training on mortality, CVD and oncologic risk have not been investigated in the NASH population. At this time, longitudinal study of the benefit of sustained exercise intervention on not only cardiovascular risk but patient-centered adverse cardiovascular events is a clear unmet need and deserves future study (Fig. 4). Additionally, there is some data to suggest that regular physical activity may result in a reduced risk of HCC [28]. Exercise training benefits multiple primary cancers, including breast, prostate and colorectal, by improving health-related quality of life, lowering rates of cancer recurrence, and most importantly, improving cancer-free and overall survival [29]. Whether exercise training has a multipronged positive effect on NASH by also decreasing HCC risk remains unknown but offers an intriguing avenue for future scientific inquiry given the link between HCC and NAFLD/NASH in the absence of cirrhosis [5, 30].

Fig. 4.

As liver fibrosis stage increases in NASH, cardiorespiratory fitness declines while rates of cardiovascular disease and hepatocellular carcinoma increase

We confirmed the findings of Krasnoff et al. [10] that histologic activity in NASH patients is associated with lower cardiorespiratory fitness and provide new evidence that NAS > 5 is associated with the poorest fitness levels. While this study is unable to determine causality, these findings are important because low cardiorespiratory fitness may be contributing to histologic activity in NASH and disease progression. Lifestyle change, with both dietary counseling and exercise training, can lead to histologic benefit when coupled with 5–10% weight loss or greater [31, 32]. Whether histologic improvement is mediated by gains in cardiorespiratory fitness is unknown as neither Promrat et al. nor Vilar-Gomez et al. measured cardiorespiratory fitness in their studies [31, 32], however, what seems more likely is that there is no direct causation and rather exercise training only triggers adaptive mechanisms that indirectly support liver regeneration. Our recent NASHFit study demonstrated that independent of weight loss, hepatic steatosis as measured by magnetic resonance imaging proton density fat fraction (MRI-PDFF), can be decreased after 20-weeks of exercise training in patients with NASH in parallel with improvement in VO_2peak, confirming similar findings by Sullivan et al. [18, 22, 33]. While neither study measured histologic change following exercise, there is a growing body of evidence that MRI-PDFF can serve as a non-invasive biomarker for histologic response, especially if a 30% or greater relative reduction is achieved [34, 35]. The NASHFit study found 40% of patients achieved this threshold of PDFF reduction that surrogates for histologic response, which is similar to rates of response for early phase drug studies that use PDFF as the primary outcome [18]. While we await the definitive study demonstrating exercise-induced histologic response independent of weight loss that also measures changes in cardiorespiratory fitness, it is reasonable to presume that gains in cardiorespiratory fitness may play a role in histologic NASH improvement because an increase in fitness level leads to normal physiologic adaptations in the liver including resistance to oxidative stress, decreased inflammation, downregulation of lipogenic enzymes and augmentation of lipid metabolism [36].

Despite being the largest study to date to explore the relationship between liver fibrosis and cardiorespiratory fitness, our study has several limitations. Subgroup analysis by individual fibrosis score (e.g., F1 vs. F2) was underpowered due to sample size and accordingly, histologic fibrosis stage F3 was chosen as the categorical cut-off value based on previous studies correlating advanced fibrosis with CVD, HCC and death. Additionally, patients with cirrhosis were excluded given the expected high fail rates of graded exercise testing and that this study population, even if well compensated and without clinically significant portal hypertension, is distinctly different than the non-cirrhotic population. This limits generalizability of our findings. Due to similar concerns related to safety of graded exercise testing, patients with BMI > 45 kg/m² were excluded. Additionally, the study population was predominantly women which could have impacted the results. Different modalities of maximal exercise testing were employed during this study (e.g., graded cycle ergometer 15 Watt incremental protocol and treadmill testing with the Bruce protocol), however, previous studies have shown that the results from these testing modalities are comparable, although not fully interchangeable as predicted VO_2peak via treadmill may be slightly higher than that obtained using cycle ergometer [37, 38]. Despite this potential difference, similar rates of maximal testing based on a priori definitions were found independent of testing modality. The selection of sedentary adults may also introduce an element of selection bias which may have led to our findings of lower VO_2peak than previously reported, although it is well known that the majority of patients with NAFLD and NASH are sedentary and do not achieve recommended amounts of weekly physical activity [39]. An additional limitation is not collecting baseline hemoglobin values as anemia, if present, can alter oxygen carrying capacity. We also did not systematically capture sarcopenia.

In conclusion, patients with NASH and advanced liver fibrosis have the poorest cardiorespiratory fitness as measured by relative VO_2peak. Patients with NASH have lower levels of cardiorespiratory fitness than previously thought. Histologic NASH activity is also associated with cardiorespiratory fitness. Further research is needed to determine if exercise training can improve cardiorespiratory fitness across multiple stages of liver fibrosis, degree of NAS and directly lessen morbidity and mortality from this common condition.

Abbreviations

AASLD

American Academy for the Study of Liver Diseases

ACSM

American College of Sports Medicine

AGA

American Gastroenterological Association

ALT

Alanine aminotransferase

AST

Aspartate aminotransferase

BMI

Body mass index

CI

Confidence interval

CV

Cardiovascular

CVD

Cardiovascular disease

EIM

Exercise is Medicine

FRIEND

Fitness Registry and the Importance of Exercise National Database

HCC

Hepatocellular carcinoma

HDL

High density lipoprotein

HR

Heart rate

IRB

Institutional Review Board

LDL

Low density lipoprotein

MET

Metabolic equivalent

NAFLD

Nonalcoholic fatty liver disease

NAS

NASH activity score

NASH

Nonalcoholic steatohepatitis

RER

Respiratory exchange ratio

RPE

Rate of perceived exertion