Search

We performed the survey investigating the relationship between autistic traits and body temperature with the preregistered protocol ( https://osf.io/z5dmj ). In our preregistered protocol, we planned to exclude participants if their AQ, EQ, SQ, Morningness/Eveningness (CSM), body temperature, and Time ([time of body temperature measurement - time of waking up]) exceed ±2 SD. After the data collection, we have modified our data analyses as follows: 1. As for the body temperature, the criterion of ±2 SD resulted in excluding the data which can be regarded as within a normal range. We thus used the criterion of ±3 SD. 2. As for AQ, EQ-SQ, and CSM, the participants regarded as Satisficer and those who marked the same value in each questionnaire were excluded. These endorsed that the estimated scores were reliable. We also noted that it should be reasonable that people with diagnostic levels show higher values for AQ and EQ-SQ. Therefore, we did not perform outlier exclusions. 3. The Time index was not normally distributed and was heavily skewed. We decided that the data is excluded if Time index exceeds 24 hours. We also planned to standardize the body temperature data if we find a significant main effect of the manufacturer of axillary thermometers or interaction effects in the analysis of variance (ANOVA) for body temperatures with factors the manufacturer and age group. Only 58% of the participants gave information about manufacturers, and only three manufacturers were identified as available for the ANOVA. We, therefore, did not standardize the temperature data of the specific manufacturers but checked whether the results differed if the specific manufacturer’s data was excluded. We excluded two participants’ data whose body temperatures were measured not by axillary thermometers based on the information of the manufacturer and model. The summary of our findings is as follows; [All data] 1. Multiple regression analysis with age, Morningness/Eveningness (CSM), Time, and their interaction terms for body temperatures showed that only age had a significant negative relationship. The input of AQ, EQ, and SQ did not change the explanatory effect of the model. 2. Multiple regression analysis with age, Time, and their interaction terms for CSM found a significant positive relationship with age. The input of AQ, EQ, and SQ increased the explanatory effect of the model, and there was a significant negative relationship for AQ. 3. Partial correlation analyses between age and AQ, EQ, or SQ showed that AQ and EQ were negatively correlated with and SQ was positively correlated with age. 4. There was a negative correlation between AQ and EQ. Also, there was a negative correlation between AQ and SQ, and a positive correlation between EQ and SQ. 5. The ANOVA was performed for body temperatures with 4 levels of thermometer manufactures (three companies over 100 reported cases and the other, unlabeled data) x age labels (the 20s-70s) showed a significant main effect for the manufacturer as well as age. The values of one company (label 2) were higher than those of the unlabeled data (label 0), and the values of another company (label 3) were lower than those of the other companies (label 1 and 2). We confirmed that the results were consistent with and without these two companies’ (labels 2 and 3) data. [Comparison between gender] 1. T-tests between gender for body temperature, AQ, EQ, SQ, Time, wake-up time, and measurement time showed that AQ and SQ were higher in men, and EQ was higher in women. [Analyses in each gender] #Male 1. The multiple regression analysis found a significant negative relationship between age and body temperature. Furthermore, the input of AQ, EQ, and SQ increased the explanatory effect of the model. AQ and EQ were negatively associated with body temperature. 2. The multiple regression analysis found a significant positive relationship between age and CSM. The input of AQ, EQ, and SQ increased the explanatory power of the model, and there was a significant negative relationship for AQ and a positive significant relationship for SQ. 3. Partial correlation analyses showed that AQ and EQ were negatively correlated with and SQ was positively correlated with age. 4. There was a negative correlation between AQ and EQ. Also, there was a negative correlation between AQ and SQ, and a positive correlation between EQ and SQ. 5. The ANOVA for the body temperature found that the values of one company (label 3) were significantly lower than those of another company (label 2) and the unlabeled data (label 0). We confirmed that the results were consistent with and without this one company’s (labels 3) data. #Female 1. The multiple regression analysis found that age had a significant negative relationship with body temperature. The input of AQ, EQ, and SQ did not increase the explanatory effect of the model. 2. The multiple regression analysis for CSM showed a significant positive relationship with age. The input of AQ, EQ, and SQ increased the explanatory power of the model, and there was a significant negative relationship for AQ. 3. Partial correlation analyses showed that AQ and EQ were negatively correlated with age. 4. There was a negative correlation between AQ and EQ. Also, there was a negative correlation between AQ and SQ, and a positive correlation between EQ and SQ. 6. The ANOVA for the body temperature found that the values of one company (label 2) were significantly higher than those of the unlabeled data (label 0). We confirmed that the results were consistent with and without this one company’s (labels 2) data. # Further comparisons between gender 1. The results of multiple regression analysis on body temperature and the correlation between SQ and age were different between the male and female data. We performed an ANOVA with age group × gender for body temperature, AQ, EQ, and SQ. As for body temperature, there was a significant interaction: Women showed lower body temperature in the 60s and 70s. The multiple regression analysis on body temperature with all data but excluding the 60s and 70s females’ ones showed the results consistent with the male data. There was no significant interaction between age group and gender for AQ, EQ, and SQ.

Supplemental Material for Association between early immunophenotypic changes and therapeutic response of belimumab in patients with systemic lupus erythematosus by Aiko Hirano, Kazuki Fujioka, Takashi Kida, Satoshi Omura, Hideaki Sofue, Aki Sakashita, Tomoya Sagawa, Yu Isoda, Akiko Kasahara, Risa Sagawa, Wataru Fujii, Takahiro Seno, Makoto Wada, Masataka Kohno, and Yutaka Kawahito in Lupus