Table 2.
Univariable and multivariable linear regression of key sociodemographic characteristics and MOS-SSS scale score
| Univariable Linear Regressions | Multivariable Linear Regression | |||
|---|---|---|---|---|
| Variables | B coefficienta (95% CI) | P-value | B coefficienta (95% CI) | P-value |
| Age at interview | ||||
| Less than 40 | Ref | Not selectedb | ||
| 40 to 49 | 4.57 (-1.78, 10.91) | 0.159 | ||
| 50 to 59 | 3.09 (-2.93, 9.12) | 0.315 | ||
| 60 or more | 2.30 (-4.85, 9.46) | 0.528 | ||
| Gender | ||||
| Male | Ref | |||
| Female | 4.58 (-0.69, 9.86) | 0.089 | Not selectedb | |
| Other | 5.04 (-11.05, 21.13) | 0.539 | ||
| Sex orientation | ||||
| Heterosexual | Ref | |||
| Homosexual | -1.07 (-5.87, 3.73) | 0.662 | ||
| Other | -2.36 (-8.55, 3.82) | 0.454 | ||
| Current relationship status | ||||
| Married/Common Law/Steady Partner | Ref | Ref | ||
| Single/Dating/Divorced/ Separated/Other | -22.25 (-26.43, -18.07) | < 0.001 | -21.23 (-25.60, -16.85) | < 0.001 |
| Key populations | ||||
| MSM only | Ref | |||
| IDU only | 2.00 (-3.65, 7.65) | 0.488 | ||
| Neither MSM nor IDU | -2.34 (-7.93, 3.24) | 0.411 | ||
| Both MSM and IDU | -9.21 (-18.13, -0.29) | 0.043 | ||
| Geographic location | ||||
| Metropolitan Vancouver Areac | Ref | Not selectedb | ||
| Outside Metropolitan Vancouver Area | − 0.039 (-5.08, 4.31) | 0.871 | ||
| Ethnicity | ||||
| Caucasian | Ref | Ref | ||
| Indigenous | 4.65 (-1.46, 10.77) | 0.136 | 7.40 (1.47, 13.32) | 0.015 |
| Asian/African/Black/Caribbean/Latin | -4.01 (-12.19, 4.17) | 0.337 | -6.07 (-13.77, 1.64) | 0.123 |
| Other | -0.70 (-9.05, 7.65) | 0.869 | 2.50 (-5.21, 10.22) | 0.525 |
| Highest education | ||||
| Less than high school | Ref | |||
| Completed high school | -1.65 (-7.70, 4.40) | 0.593 | ||
| Completed greater than high school | -4.38 (-9.88, 1.11) | 0.118 | ||
| Employment status | ||||
| Working | Ref | Not selectedb | ||
| Unemployed | -5.04 (-9.96, -0.13) | 0.044 | ||
| Other | -1.26 (-6.90, 4.38) | 0.662 | ||
| Incarceration ever | ||||
|
No Yes |
Ref 1.08 (-3.45, 5.62) |
0.640 | Not selectedb | |
| Homelessness in the last 12 months | ||||
|
No Yes |
Ref -6.35 (-12.64, -0.05) |
0.048 | Not selectedb | |
| Housing stability d | ||||
| Stable | Ref | |||
| Neutral | -8.26 (-15.89, -0.63) | 0.034 | ||
| Not stable | -18.43 (-24.47, -12.38) | < 0.001 | ||
| Mental health diagnosis ever | ||||
|
No Yes |
Ref -7.03 (-11.56, -2.50) |
0.002 |
Ref -4.17 (-8.47, 0.13) |
0.057 |
| Recent hazardous alcohol use | ||||
| No | Ref | |||
| Yes | 3.61 (-0.77, 8.00) | 0.106 | ||
| Lifetime violence (n = 594) | ||||
|
No Yes |
Ref -6.22 (-11.40, -1.05) |
0.018 |
Ref -5.73 (-10.74, -0.73) |
0.025 |
| Sexual activity in the last 12 months (n = 590) | ||||
|
No Yes |
Ref 8.25 (3.75, 12.75) |
0.000 |
Ref 4.84 (0.42, 9.26) |
0.032 |
| History of IDU | ||||
| Never | Ref | Ref | ||
| Yes but more than a year ago | 2.69 (-2.67, 8.05) | 0.326 | 6.58 (1.43, 11.73) | 0.012 |
| Yes in the past year | -6.68 (-12.21, -1.15) | 0.018 | -2.07 (-7.47, 3.34) | 0.453 |
Bold text refers to p-values < 0.05
CI: confidence interval
MSM: men who have sex with men
IDU: injection drug use
aB coefficient is used to represent the unstandardized beta coefficient
bExplanatory variables were selected for potential inclusion in the multivariable model based on univariable associations and a priori knowledge from previous literature and consultation with living experience. Model selections were conducted using a backward stepwise technique which is based on Type III p-values and Akaike information criterion (AIC). The variable with the highest Type III p-value was dropped at each step of the selection process until the model reached the lowest AIC, with a lower AIC indicating better model fit
cMetropolitan Vancouver Area includes Vancouver, the largest metropolitan area in BC and the Vancouver Coastal and Fraser Health Authorities
dHousing stability is defined as whether an individual feels their housing situation is stable (i.e. housing is safe, secure and affordable)