4.1 Comparison of general demographic data between follow-up and lost follow up population
Of the 620 cognitively normal elderly subjects, 145 were lost during the 2-year follow-up, with a loss rate of 23.4%. Compared with the lost follow up participants, follow up participants were more likely to drink tea, combined with a higher proportion of hypertension, diabetes, coronary heart disease, hyperlipidemia, cerebral infarction, depression and their MOCA scores were higher (p<0.05). However, there were no statistically significant differences (p>0.05) in age, education, gender, smoker, drinker, take exercise, hobby, cerebral hemorrhage, eat rice weekly, eat bread weekly, eat fruits weekly, eat orange vegetables weekly, eat green vegetables weekly, eat fish weekly, eat tofu weekly, eat white meat weekly and eat red meat weekly between the two groups (Supplementary 1).
4.2 Comparison of general demographic data between high white rice consumption and low white rice consumpition
In the present study, we took the 14 times a week as the cut-off value to divide the low frequency and high frequency20, and the elderly with the low white rice consumption were younger, less willing to eat fish and green vegetables, and had a higher MCI conversion rate (p<0.05), while there was no statistically significant differences (p>0.05) in education, gender, smoker, drinker, tea drinker, take exercise, hobby, hypertension, diabetes, coronary heart disease, hyperlipidemia, cerebral infarction, depression, cerebral hemorrhage, consumption of bread, fruits, orange vegetables, tofu, white meat and red meat between the two groups. Table 1 presents the results.
4.3 Comparison of general demographic data between the NC-MCI group and the NC-NC group
Of the 471 older adults who were followed, 91 converted to MCI, while 380 remained cognitively normal, resulting in a 2-year incidence of MCI of 19.3%. Compared with those who had transitioned to MCI, those who had not transitioned were younger, had fewer drinks, ate less white rice per week, but were more educated and had higher MOCA scores (p<0.05), however, there were no statistically significant differences (p>0.05) in gender, smoker, tea drinker, take exercise, hobby, hypertension, diabetes, coronary heart disease, hyperlipidemia, cerebral infarction, depression, cerebral hemorrhage, eat bread weekly, eat fruits weekly, eat orange vegetables weekly, eat green vegetables weekly, eat fish weekly, eat tofu weekly, eat white meat weekly and eat red meat weekly between the two groups. Table 2 presents the results.
4.4 The results of Cox regression analysis
Table 3 shows the results of Cox regression analysis(change to MCI as a dependent variable, change time as a time variable). Higher weekly rice consumption was a risk factors for MCI (model 1: p=0.019, HR=1.051, 95%CI:1.008~1.096); This relationship remained statistically significant after adjusting for age and education in model 2: (p=0.041, HR=1.044, 95%CI:1.002~1.089); Further adjustment for drinking in model 3 also did not change the results (p=0.020, HR=1.052, 95%CI:1.008~1.097). The ROC curve was used to determine the weekly rice consumption to predict the risk of MCI, and it was found that the area under the curve was 0.591 (p=0.07, 95%CI:0.527~0.655), suggesting that the weekly rice consumption had a mild-to-moderate effect in predicting MCI. Figure 2 presents the results.