Abstract: Background：Mild neonatal hypoxic-ischemic encephalopathy (HIE) has been considered to have better long-term outcomes. In recent years, there was evidence that the long-term outcomes of mild HIE were worse than those of normal children. Objective：To explore whether mild neonatal HIE benefits from hypothermia therapy. Design：Systematic review and meta-analysis. Methods：Literature was searched in databases of PubMed, Embase, Cochrane, CINAHL from the establishment to Dec 6,2021, and in database of SinoMed from the establishment to Nov 12, 2021. Retrieved papers were first screened by titles and abstracts. A second phase of screening was subsequently undertaken to screen papers for exclusion criteria as follows either: a. the diagnosis of mild neonatal HIE failing to meet the modified Sarnat standard; b. neonates with congenital malformations; c. failure to extract the outcome data of neonatal mild HIE; d. the follow-up time <12 months. GRADE methodology was used to rate the quality of evidence body. Publication bias analysis was adopted for more than 10 articles in either evidence body, and extracted data were analyzed using Revman 5.4 and R. The I2 test was used for heterogeneity. If I2≤50%, the fixed effect model should be adopted; if I2>50%, the random effect model should be adopted. Main outcome measures：Rate of death and major neurodevelopmental disabilities at more than 12 months. Results：1,839 articles were identified through database searching. A total of 113 full-text articles were assessed for eligibility and 24 articles were included in this review. The review was conducted for 4 RCTs, 7 cohort studies, and 13 case series reports. A total of 21 literature reported nervous system disability by dichotomous variables, and three literature reported nervous system disability by continuous ones. Follow-up time in 16 articles was from 12 months to 3 years, and 9 articles over 3 years (1 article reported neurological disability at 24 months and 7 years old respectively). No deaths were reported in the 24 articles during the observation period. A total of 369 infants with mild neonatal HIE was reported in 21 articles with dichotomous variables and the results showed that the incidence of major neurodevelopmental disabilities in mild HIE survivors was 21% (95%CI: 0.14-0.29), I2=80%, and there was publication bias (Egger test, t=4.68, P<0.01). The incidence of major neurodevelopmental disabilities was 11% (95%CI: 0.00-0.23) on therapeutic hypothermia (6 articles) and 21 percent (95%CI: 0.13-0.29) on non-therapeutic hypothermia (19 articles). There was no difference (OR=0.78, 95%CI: 0.27-2.31, I2=0) in the incidence of major neurodevelopmental disabilities between mild HIE patients treated with therapeutic hypothermia and those not treated with therapeutic hypothermia (4 RCTs), and patients who received non-therapeutic hypothermia compared with normal children (4 cohort studies) had higher incidence of major neurodevelopmental disabilities (OR=19.06, 95%CI: 7.01-51.85, I2=42%). The incidence of neurological disability was 20 percent (95%CI: 11%-29%) in subgroups whose follow-up time was from 12 months to 3 years and 24 percent (95%CI: 11%-36%) in over 3 years. Conclusion：When patients diagnosed with mild HIE was followed up until the age of 3 years old, the potential incidence of major neurodevelopmental disabilities increased to 24%, and benefits of combining with support treatment of therapeutic hypothermia as auxiliary is greater than the disadvantages.

Key words: Neonate, Mild , Hypoxic ischemic encephalopathy, Therapeutic hypothermia, Nervous system disability