Meta-Analysis
Copyright ©The Author(s) 2015.
World J Gastroenterol. Feb 7, 2015; 21(5): 1650-1662
Published online Feb 7, 2015. doi: 10.3748/wjg.v21.i5.1650
Table 1 Baseline characteristics for studies included in meta-analysis
First author and yearCountryCase, n (M/W)Control, n (M/W)DesignExcluding1Cut-offAdjusted factorsNOS
High vs low category in waist circumference
Zelber-Sagi et al[27]Israel66/32108/120c-sHep B, C, DILI102/8858
Park et al[28]Korea62/84118/86c-sHep B, C90/80116
Choi et al[29]Korea246/68216/129c-sHep B, C90/8076
Tsai et al[30]Taiwan253/120275/228c-sNone90/8077
Kogiso et al[31]Japan33/2460/113c-sHep B, C85/90107
Das et al[32]India89/75926/821n-c-cHep B, C, DILI90/8068
Pinidiyapathirage et al[33]Sri Lanka45/27144/185c-sHep B, C90/8048
High vs low category in waist-to-hip ratio
Kogiso et al[31]Japan33/2460/113c-sHep B, C0.845/0.845107
Zheng et al[34]China189/61192/48c-cHep B, C0.900/0.85087
Wengert et al[35]Germany58/22138/125c-sHep B, CNR89
Per-unit increase in waist circumference
Yoon et al[36]Korea11/2715/38c-sOther liver diseases/37
Church et al[22]USA24/0218/0n-c-cOther liver diseases/37
Sung et al[37]Korea7155/186710461/11757c-sHep B, C, DILI/26
Seo et al[20]Korea45/2850/36c-cHep B, C, DILI/128
Xu et al[38]China139/88416/235c-sOther liver diseases/77
Sathiaraj et al[39]India77/2175/27c-cHep B, C, DILI/26
Zhou et al[40]China211513CohortOther liver diseases/118
Eshraghian et al[21]Iran127705c-sDILI, other liver diseases/127
Cheah et al[41]Malaysia17/1719/24c-sNone/98
Foster et al[42]USA227/2941144/1391c-sDILI, other liver diseases/97
Li et al[23]China28/51130/233CohortOther fatty liver/178
Table 2 Subgroup and meta-regression analyses of waist circumference and the risk of nonalcoholic fatty liver disease
CovariatesSubgroupStudies, nOR(95%CI)P of QI2Inter-groupP valueUnivariateP valueMultivariateP value
Per-unit increase in waist circumference
RegionEast91.076 (1.025-1.129)0.00073.60.0060.1710.214
West21.031 (1.016-1.046)0.5350.0
Designc-s81.063 (1.014-1.114)0.00079.10.1570.7640.984
Others31.071 (1.026-1.119)0.25426.9
AdjustedYes71.051 (1.019-1.084)0.03555.80.0310.4890.770
(> 3 confounders)No41.126 (1.000-1.269)0.00085.0
IndividualsYes61.054 (1.025-1.083)0.05556.70.1630.7280.975
(> 700)No51.100 (0.991-1.22)0.00081.5
High vs low category of waist circumference
Designc-s52.113 (1.615-2.765)0.18036.20.0570.6420.947
Others24.072 (2.189-7.575)0.5120.0
IndividualsYes32.070 (1.546-2.774)0.24528.90.1220.6350.942
(> 600)No43.185 (2.010-5.046)0.16441.2
Table 3 Further analyses after adding several studies
CovariatesSubgroupn of studiesOR(95%CI)P of QI2Inter-studyP valueUnivariateP valueMultivariateP value
Per-unit increase in waist circumference
RegionEast151.089 (1.057-1.122)0.00094.10.0010.1800.145
West51.040 (1.028-1.052)0.14940.8
Designc-s141.067 (1.042-1.193)0.00094.60.0100.7340.968
Others61.076 (1.049-1.103)0.14039.8
AdjustedYes111.059 (1.038-1.081)0.00082.90.8060.393/
(> 3 factors)No91.101 (1.055-1149)0.00095.9
Adjusting for BMIYes141.064 (1.040-1.089)0.00084.80.0000.6490.918
No61.086 (1.046-1.127)0.00095.3
Excluding ALDYes161.079 (1.051-1.107)0.00090.00.0000.6030.869
No41.056 (1.022-1.093)0.00093.8
High vs low category of waist circumference
RegionEast112.687 (2.018-3.579)0.00074.40.0000.5450.934
West53.098 (2.005-4.785)0.00083.3
Designc-s122.471 (1.958-3.119)0.00072.30.0000.0150.078
Other45.212 (2.121-12.807)0.8580.0
AdjustedYes103.002 (2.268-3.973)0.00073.10.0000.459/
(> 3 factors)No62.840 (2.184-3.691)0.00075.2
Adjusting for BMIYes102.492 (1.898-3.271)0.00265.80.0000.2210.591
No63.281 (2.308-4.664)0.00079.5
Excluding ALDYes122.856 (2.115-3.855)0.00074.90.0000.9101.000
No42.794 (1.716-4.549)0.00089.5
Table 4 Summary estimates by fixed- vs random-effects methods
OR (95%CI)
Fixed-effectsRandom-effects
Categorical variable
WC2.344 (1.831-3.000)2.550 (1.799-3.615)
WHR3.910 (2.255-6.780)4.061 (1.529-10.790)
BMI2.183 (1.582-3.013)2.854 (1.604-5.080)
WC12.844 (2.082-3.885)3.139 (2.067-4.767)
Per-unit increase
WC1.043 (1.031-1.055)1.065 (1.029-1.103)
BMI1.121 (1.093-1.150)1.250 (1.131-1.382)