Research Article
Time-Series Analysis of Continuously Monitored Blood Glucose: The Impacts of Geographic and Daily Lifestyle Factors
Table 5
Frequency of variable significance in 5-minute ARIMA models.
| Variables | All day models |
Waking hours only models |
| At-home time | 11 | (33%) | 6 | (18%) | Automobile time | 9 | (27%) | 10 | (30%) | Public transit time | 6 | (18%) | 5 | (15%) | Shopping time | 5 | (15%) | 4 | (12%) | Calories consumed | 4 | (12%) | 7 | (21%) | Distance home | 4 | (12%) | 6 | (18%) | Exercise time | 2 | (6%) | 2 | (6%) | Meds taken (yes/no) | 2 | (6%) | 1 | (3%) | Time spent with others | 2 | (6%) | 3 | (9%) | Restaurants time | 1 | (3%) | 2 | (6%) | Insulin taken (yes/no) | 0 | (0%) | 1 | (3%) | Movement intensity | 0 | (0%) | 0 | (0%) | Trip time | 0 | (0%) | 0 | (0%) | Work/school time | 0 | (0%) | 0 | (0%) | Out-of-home time | 0 | (0%) | 0 | (0%) |
|
|