Skip to main content
. 2014 Jul 22;11(7):e1001680. doi: 10.1371/journal.pmed.1001680

Table 4. Adjusted effect estimates for non-alcoholic fatty liver disease, non-alcoholic steato-hepatitis, advanced (stage F3) fibrosis and prevalent/incident chronic kidney disease, based on individual participant data meta-analysis from 20 studies (29,282 participants).

Outcome Cross-sectional Studies Longitudinal Studies
Covariate Participants OR (95% CI) p-Value Participants HR (95% CI) p-Value
CKD in NAFLD vs. non-NAFLD
Diabetes 507 diabetic 3,031 non-diabetic 2.62 (1.95–3.57) 0.00001 2,046 diabetic, 39,365 non-diabetic 1.99 (1.83–2.25) 0.00001
Age (y) a 51 (18–89) 2.39 (1.93–2.99) 0.00003 38 (20–80) 2.10 (1.74–2.63) 0.000006
BMI (kg/m2) a 26 (16–71) 2.21 (1.85–2.67) 0.00006 24 (16–47) 2.40 (1.94–2.99) 0.00001
Met sy 885 with Met Sy, 2,654 without Met Sy 2.43 (1.95–3.06) 0.00001 3,758 with Met Sy, 19,226 without Met Sy 2.13 (1.77–2.71) 0.00009
HTN 1,061 with HTN, 2,477 without HTN) 2.19 (1.78–2.61) 0.00002 4,367 with HTN, 18,617 without HTN 2.01 (1.69–2.77) 0.00002
Smoking 1,168 smokers, 2,370 non-smokers 2.11 (1.75–2.58) 0.00005 9,653 smokers, 13,331 non-smokers 2.28 (1.87–2.89) 0.00008
Ethnicity 2,867 Asians, 1,071 non-Asian 2.25 (1.87–2.70) 0.00002 18,519 Asians, 6,937 non-Asians 2.32 (1.32–4.10) 0.00004
Cirrhosis 2,976 non-cirrhotic, 562 cirrhoticb 2.34 (1.92–2.85) 0.00001 No cirrhotic participant
Waist (cm) a 91 (54–153) 2.63 (2.02–3.12) 0.00007 89 (51–150) 2.52 (2.00–3.02) 0.00007
HOMA-IR a 1.8 (1.0–15.1) 2.55 (2.00–2.98) 0.00001 1.6 (0.6–10.3) 2.12 (1.76–2.54) 0.0009
Duration of follow-up a 5 (1–29) 2.48 (1.98–2.97) 0.0001
All covariates 3,538 individuals 1.95 (1.55–2.71) 0.00001 22,984 individuals 1.91 (1.68–2.21 0.00001
CKD in NASH vs. simple steatosis
Diabetes 119 diabetic, 769 non-diabetic 2.77 (1.81–4.24) 0.0004 96 diabetic, 333 non-diabetic 2.32 (1.55–3.48) 0.00009
Age (y) a 46 (18–80) 3.01 (2.50–3.72) 0.00001 47 (18–67) 2.61 (1.71–3.24) 0.00004
BMI (kg/m2) a 33 (18–59) 2.78 (2.09–3.24) 0.0003 27 (18–47) 2.32 (1.59–3.15) 0.00006
Met sy 355 with Mey Sy, 532 without Met Sy 2.66 (2.05–3.12) 0.00008 136 with Met Sy, 293 without Met Sy 2.71 (2.18–3.59) 0.0001
HTN 408 with HTN, 479 without HTN 2.59 (1.97–3.18) 0.00002 172 with HTN, 257 without HTN 2.58 (1.99–3.11) 0.00007
Smoking 222 smokers, 665 non-smokers 2.69 (1.63–3.82) 0.00001 103 smokers, 326 non-smokers 2.56 (1.54–3.13) 0.0002
Ethnicity 263 Asians, 624 non-Asians 3.14 (2.44–4.96) 0.00001 102 Asians, 327 non-Asians 2.37 (1.41–3.98) 0.0001
Cirrhosis No cirrhotic participant No cirrhotic participant
Waist (cm) a 110 (51–162) 2.62 (2.01–3.31) 0.0006 99 (70–115) 2.76 (2.01–3.48) 0.0003
HOMA-IR a 3.1 (0.3–28.1) 2.54 (1.98–2.96) 0.00002 2.7 (0.8–3.9) 2.24 (1.55–2.46) 0.00001
Duration of follow-up a 14 (3–30) 2.31 (1.42–2.87) 0.00007
All covariates 887 individuals 2.42 (1.80–3.84) 0.0001 429 individuals 2.01 (1.40–2.98) 0.0001
CKD in advanced (stage F3) vs. non-advanced (stage F0–F2) fibrosis
Diabetes 119 diabetic, 769 non-diabetic 5.39 (3.66–8.20) 0.0001 96 diabetic, 333 non-diabetic 3.88 (2.97–5.23) 0.00001
Age (y) a 46 (18–80) 5.12 (3.71–6.99) 0.00009 47 (18–67) 3.91 (2.56–4.98) 0.00007
BMI (kg/m2) a 33 (18–59) 4.89 (3.94–5.98) 0.00001 27 (18–47) 3.45 (2.51–4.02) 0.00004
Met sy 355 with Mey Sy, 532 without Met Sy 5.00 (4.11–5.97) 0.00007 136 with Met Sy, 293 without Met Sy 3.28 (2.39–4.52) 0.00006
HTN 408 with HTN, 479 without HTN 4.98 (3.81–6.02) 0.0003 172 with HTN, 257 without HTN 4.11 (2.88–5.51) 0.0001
Smoking 222 smokers, 665 non-smokers 5.27 (3.94–6.42) 0.00008 103 smokers, 326 non-smokers 3.29 (2.24–4.43) 0.00002
Ethnicity 263 Asians, 624 non-Asians 5.18 (3.57–6.82) 0.00001 102 Asians, 327 non-Asians 3.70 (1.46–9.39) 0.0006
Cirrhosis No cirrhotic participant No cirrhotic participant
Waist (cm) a 110 (51–162) 5.39 (4.21–6.13) 0.0006 99 (70–115) 3.39 (2.12–4.51) 0.00001
HOMA-IR a 3.1 (0.3–28.1) 5.13 (4.01–6.29) 0.00004 2.7 (0.8–3.9) 3.70 (2.12–4.91) 0.00003
Duration of follow-up a 14 (3–30) 3.58 (2.46–5.89) 0.0009
All covariates 887 individuals 4.86 (3.54–6.69) 0.00001 429 participants 3.00 (2.08–4.33) 0.0001

Data from all studies providing IPD were pooled together into a single dataset and effect estimates were calculated using multivariate logistic regression (cross-sectional studies) or Cox proportional hazard models (longitudinal studies). In these models, studies were incorporated as cluster and treated as random-effect, while covariates were treated as fixed-effect. The individual patient covariates entered in the models were: age, BMI, metabolic syndrome, hypertension, smoking status, diabetes, ethnicity (Asian versus non-Asian population), presence of cirrhosis, waist circumference, HOMA-IR index, duration of follow-up (for longitudinal studies). Finally, a fully adjusted model was run, with all covariates entered.

HTN, hypertension; Met Sy, metabolic syndrome; waist, waist circumference.

a

For continuous variables, median (range) of values is reported.

b

All cirrhotic individuals derive from the study by Park et al. [30].