Table 2 The strength of epidemiologic evidence of 22 meta-analyses of observational studies that affect the risk of NAFLD
Exposure | Measure | Reference | Precision of the estimate | Consistency of results | No evidence of small-study effects | Grade | |
|---|---|---|---|---|---|---|---|
>1000 disease cases | P < 0.001 | I2 < 50% and Cochran Q test P > .10 | P > 0.1 | ||||
15 factors that increased the risk of NAFLD | |||||||
 Smoking | NAFLD | Rezayat2017 | Yes | No | Yes | Yes | Low |
 Passive smoking | NAFLD | Rezayat2017 | No | No | No | Yes | Low |
 Former smoking | NAFLD | Rezayat2017 | No | No | Yes | Yes | Low |
 Soft drinks | NAFLD | He2020 | Yes | Yes | Yes | No | Moderate |
 sugar sweetened beverages | NAFLD | Asgar-Taee2018 | Yes | No | Yes | Yes | Low |
 Sugar-Sweetened Soda | NAFLD | Wijarnpreecha2015 | Yes | Yes | Yes | Yes | High |
 Hypercaloric fructose diet | IHLC | Chung2014 | No | Yes | Yes | Yes | Moderate |
 Red meat | NAFLD | He2020 | Yes | No | No | No | Low |
 Short sleep | NAFLD | Wijarnpreecha2016 | Yes | No | Yes | Yes | Low |
 Obesity | NAFLD | Li2016 | Yes | Yes | No | Yes | Moderate |
 Per 1-unit increase in WC | NAFLD | Pang2015 | Yes | Yes | No | Yes | Moderate |
 Per 1-unit increase in BMI | NAFLD | Pang2015 | Yes | Yes | No | Yes | Moderate |
 WHR | NAFLD | Pang2015 | No | No | No | Yes | Low |
 Hyperuricemia | NAFLD | Darmawan2017 | Yes | Yes | No | Yes | Moderate |
 Hyperuricemia | NAS | Jaruvongvanich2017 | No | Yes | Yes | Yes | Moderate |
7 factors that decreased the risk of NAFLD | |||||||
 Modest alcohol | NAFLD | Sookoian2014 | Yes | Yes | No | Yes | Moderate |
 Modest alcohol | NASH | Sookoian2014 | No | Yes | Yes | Yes | Moderate |
 Coffee | NAFLD | Chen2018 | Yes | Yes | No | Yes | Moderate |
 Coffee | Liver fibrosis | Wijarnpreecha2017 | No | Yes | No | No | Low |
 Green tea | Liver steatosis | Yin2015 | No | No | No | Yes | Low |
 Nut | NAFLD | He2020 | Yes | No | Yes | Yes | Low |
 Weight loss | NASH | Koutoukidis2019 | No | No | Yes | No | Low |