Table 2.
Network characteristics
| Characteristic | Asymptomatic, n = 1378 | MS, n = 115 | P-valuea |
|---|---|---|---|
| Network structure b | |||
| Size, median [IQR] | 8.00 [6.00, 12.00] | 8.00 [5.00, 11.50] | 0.130 |
| Density, median [IQR] | 67.00 [50.00, 89.00] | 69.00 [53.00, 90.00] | 0.170 |
| Constraint, median [IQR] | 44.00 [37.72, 53.03] | 44.71 [38.19, 56.17] | 0.315 |
| Effective size, median [IQR] | 4.00 [2.80, 5.25] | 3.55 [2.50, 5.07] | 0.053 |
| Maximum degree, median [IQR] | 5.00 [4.00, 7.00] | 5.00 [4.00, 8.00] | 0.987 |
| Mean degree, median [IQR] | 4.00 [2.80, 5.00] | 4.00 [2.50, 5.40] | 0.493 |
| Network Composition–General c | |||
| Percent kin, median [IQR] | 43 [30, 62] | 50 [33,67] | 0.205 |
| Percent who are supportive, median [IQR] | 38 [25, 50] | 40 [21,50] | 0.561 |
| Standard deviation of age, median [IQR] | 12.76 [10.04, 15.38] | 12.98 [10.54, 16.89] | 0.161 |
| Diversity of sex, median [IQR] | 0.89 [0.64, 0.96] | 0.82 [0.64, 0.96] | 0.108 |
| Diversity of race, median, Percentile {90th, 95th, 99th,100th}d | 0 {0.44, 0.55, 0.72, 1.20} | 0 {0.41, 0.59, 0.77, 0.77} | 0.046 |
| Percent contacted weekly or less often, median [IQR] | 67 [50, 80] | 67 [45, 80] | 0.896 |
| Percent who have been known for less than 6 years, median [IQR] | 20 [0, 43] | 12 [0, 33] | 0.001 |
| Percent who live more than 15 miles away, median [IQR] | 33 [17, 50] | 33 [20, 56] | 0.514 |
| Network Composition–Health Habits e | |||
| Percent who smoke, median [IQR] | 0 [0, 20] | 0 [0, 40] | 0.164 |
| Percent who do not exercise, median [IQR] | 33 [14, 54] | 25 [10, 50] | 0.068 |
| Percent who do not take medications regularly, median, Percentile {90th, 95th, 99th,100th} | 0 {0, 14, 33, 100} | 0 {0, 17, 24, 50} | 0.709 |
| Percent who do not go to doctor’s appointments, median, Percentile {90th, 95th, 99th,100th} | 0 {0, 12, 25, 100} | 0 {0, 15, 48, 100} | 0.314 |
| Percent who have a negative influence on health, median, Percentile {90th, 95th, 99th,100th} | 0 {29, 46, 71, 100} | 0 {20, 33, 78, 100} | 0.150 |
aP-values calculated using Wilcoxon signed-rank test
bNetwork structure is quantified into graph theoretic statistics. See definitions in Methods
cNetwork composition–General is the range of characteristics of people around the participant. See definitions in Methods
dPercentile is used to better understand the right-skewed distribution of the variables of race and certain health habits
eNetwork composition–Health Behavior is the range of health habits of people around the participant