Table 2.
Summary of bivariate and multivariable analyses when loneliness is the outcome (as measured by UCLA-3 Total scores)
| Variable |
Risk Direction (Lonelier) |
Bivariate Analyses |
Multivariable Analyses: Final AIC Model |
||||
|---|---|---|---|---|---|---|---|
| Statistic | value | p-value | B | 95% CI | p-value | ||
| CRPa | – | ρ | 0.12 | 0.278 | – | – | – |
| IL-6a | – | ρ | 0.10 | 0.355 | – | – | – |
| CCL2/MCP-1a | – | ρ | 0.06 | 0.568 | 0.20 | [0.11, 2.56] | 0.029 |
| sCD14a | – | ρ | 0.14 | 0.211 | 0.19 | [0.09, 1.84] | 0.033 |
| D-dimera | Higher | ρ | 0.26 | 0.017 | 0.32 | [0.21, 2.0] | < 0.001 |
| Age | Younger | ρ | -0.12 | 0.294 | -0.24 | [-0.36, 0.41] | 0.083 |
| Social Interactions | Fewer | ρ | 0.26 | 0.016 | – | – | – |
| Emotional Support | Less often | t | 4.25 | < 0.001 | 0.45 | [0.38, 1.62] | < 0.001 |
| Race/ethnicity | |||||||
| White, non-Hispanic | – | t | 0.14 | 0.887 | – | – | – |
| Hispanic | Non-Hispanic | t | 2.24 | 0.028 | – | – | – |
| Black | – | t | 0.82 | 0.416 | – | – | – |
| Otherb | Otherb | t | 2.43 | 0.018 | – | – | – |
| Nadir CD4+ T-cell countc | Higher | ρ | 0.30 | 0.005 | 0.35 | [0.10, 1.87] | < 0.001 |
| Lifetime substance use disorderd | Present | t | 2.15 | 0.034 | 0.24 | [0.06, 2.68] | 0.007 |
| Model Statistics | |||||||
| F-ratio | 10.3 | ||||||
| df | (7,74) | ||||||
| R2 | 0.49 | ||||||
| p-value | < 0.001 | ||||||
In multivariable analyses, the final model was established using backwards stepwise regression using the Akaike information criterion (AIC); UCLA-3 = UCLA Loneliness Scale-Version 3; CRP = C-Reactive Protein; IL-6 = Interleukin-6; CCL2/MCP-1 = Chemokine (C–C motif) ligand 2/monocyte chemoattractant protein-1; sCD14 = soluble cluster of differentiation 14
alog transformed; bSelf-reported race/ethnicity as other than White, Black, or Hispanic (e.g., more than one race); csquare root transformed; dLifetime abuse or dependence on substances other than alcohol or cannabis