Table 3.
Independent variable | Univariable model | Multivariable model | Final model |
OR (95% CI), P value | OR (95% CI), P value | OR (95% CI), P value | |
Age (continuous) | 1.02 (1.00–1.03), P = .050 | 1.01 (0.99–1.03), P = .457 | 1.01 (0.99–1.03), P = .310 |
Sex, female vs male | 0.54 (0.37–0.78), P = .001 | 0.60 (0.39–0.92), P = .019 | 0.59 (0.38–0.90), P = .016 |
Race/ethnicity, white vs other | 0.66 (0.16–1.90), P = .500 | – | – |
Alcohol consumption, vs never | – | ||
Once per mo | 1.57 (0.99–2.50), P = .056 | 1.23 (0.73–2.06), P = .441 | |
2 to 4 times per mo | 2.43 (1.48–3.98), P < .001 | 1.54 (0.86–2.76), P = .142 | |
4 to 6 times per mo | 3.14 (1.62–5.88), P < .001 | 1.49 (0.66–3.21), P = .317 | |
More than 6 times per mo | 2.52 (0.55–8.46), P = .170 | 1.60 (0.29–6.65), P = .542 | |
Unknown | 1.00 (0.29–2.65), P = .996 | 0.53 (0.13–1.68), P = .323 | |
Smoking status, vs non-smoker | |||
Ex-smoker | 1.31 (0.86–1.99), P = 2.00 | 1.20 (0.74–1.96), P = .456 | 1.28 (0.79–2.04), P = .311 |
Current smoker | 2.29 (1.38–3.76), P = .001 | 2.01 (1.09–3.66), P = .023 | 2.31 (1.30–4.03), P = .004 |
Unknown | 1.63 (0.71–3.38), P = .212 | 2.31 (0.89–5.52), P = .069 | 1.92 (0.77–4.40), P = .138 |
Employment, vs full-time | |||
Part-time | 2.03 (1.22–3.34), P = .005 | 1.80 (1.02–3.13), P = .039 | 1.75 (1.00–3.01), P = .046 |
Self-employed | 1.45 (0.71–2.79), P = .283 | 0.91 (0.40–1.94), P = .821 | 0.92 (0.41–1.93), P = .830 |
Student | 1.79 (0.41–5.64), P = .368 | 1.93 (0.39–7.08), P = .360 | 1.83 (0.38–6.59), P = .397 |
Retired | 3.05 (1.72–5.31), P < .001 | 2.22 (1.07–4.51), P = .030 | 2.06 (1.01–4.14), P = .045 |
Unemployed | 2.47 (1.28–4.55), P = .005 | 2.32 (1.12–4.66), P = .020 | 2.08 (1.02–4.11), P = .039 |
Homemaker | 2.12 (0.87–4.64), P = .074 | 1.96 (0.73–4.79), P = .157 | 2.07 (0.78–5.00), P = .120 |
Physically unable to work due to NASH or related complications | 9.56 (1.73–52.97), P = .007 | 28.76 (3.37–273.34), P = .002 | 26.63 (3.44–238.94), P = .002∗ |
Physically unable to work due to other reason(s) | 1.19 (0.06–6.72), P = .868 | 0.71 (0.03–6.52), P = .799 | 0.54 (0.02–4.46), P = .629 |
Years since diagnosis (continuous) | 1.18 (1.13–1.24), P < .001 | 1.15 (1.09–1.21), P < .001 | 1.17 (1.11–1.23), P < .001 |
Liver biopsy, yes vs no | 1.30 (0.91–1.90), P = .159 | 1.52 (1.00–2.33), P = .053 | 1.49 (0.98–2.28), P = .062 |
Fibrosis stage at diagnosis, vs F0 | |||
F1 | 1.26 (0.78–2.07), P = .356 | 0.99 (0.58–1.71), P = .962 | 1.00 (0.58–1.72), P = .987 |
F2 | 1.22 (0.73–2.05), P = .454 | 0.85 (0.47–1.54), P = .592 | 0.87 (0.49–1.57), P = .647 |
F3 | 0.71 (0.36–1.35), P = .304 | 0.39 (0.17–0.82), P = .016 | 0.38 (0.17–0.80), P = .014 |
F4CC | 0.09 (0.01–0.31), P = .001 | 0.06 (0.01–0.23), P = .001 | 0.06 (0.01–0.23), P < .001 |
Comorbidities | |||
Obesity from calculated BMI, yes vs no | 1.07 (0.75–1.55), P = .700 | – | – |
Obesity from physician notes, yes vs no | 2.22 (1.55–3.20), P < .001 | 1.78 (1.15–2.76), P = .010 | 1.89 (1.26–2.86), P = .002 |
Type 2 diabetes mellitus, yes vs no | 1.26 (0.84–1.86), P = .262 | – | – |
Cardiovascular disease, yes vs no | 2.26 (0.93–4.98), P = .055 | – | – |
Hypertension, yes vs no | 1.58 (1.08–2.28), P = .016 | 0.99 (0.60–1.62), P = .973 | – |
Dyslipidemia, yes vs no | 1.51 (1.04–2.18), P = .027 | 0.95 (0.57–1.58), P = .853 | – |
Non-pharmacological NASH treatment proposed at diagnosis | |||
Lifestyle change, yes vs no | 1.49 (1.04–2.15), P = .031 | 1.12 (0.71–1.76), P = .619 | – |
Behavioral strategies, yes vs no | 0.64 (0.15–1.83), P = .461 | – | – |
Bariatric or intragastric surgery, yes vs no | 1.51 (0.82–2.64), P = .159 | – | – |
Liver transplantation, yes vs no | 3.33 (1.46–7.17), P = .003 | 3.96 (1.35–11.40), P = .011 | 4.95 (1.77–13.55), P = .002 |
Pharmacological NASH treatment† | |||
Lipid-lowering drugs, yes vs no | 1.42 (0.89–2.21), P = .133 | – | – |
Vitamin E, yes vs no | 0.52 (0.27–0.94), P = .040 | 0.52 (0.25–1.01), P = .063 | 0.54 (0.27–1.02), P = .070 |
Metformin, yes vs no | 1.25 (0.73–2.06), P = .400 | – | – |
Thiazolidinediones, yes vs no | 3.95 (1.94–7.79), P < .001 | 2.03 (0.78–5.06), P = .135 | – |
Other anti-diabetic medication(s), yes vs no | 1.00 (0.23–3.01), P = .998 | – | – |
Other medication(s), yes vs no | 1.68 (0.91–2.94), P = .081 | – | – |
BMI = body mass index, CC = compensated cirrhosis, CI = confidence interval, NASH = nonalcoholic steatohepatitis, OR = odds ratio.
Only 6 patients reported inability to work due to NASH; this odds ratio should be interpreted with caution.
Pharmacological treatment prescribed at diagnosis and receiving it at date of last consultation.
Weight loss in the past 12 months, expressed as a percentage, was not significant in the univariable model (All P > .05).