# Table 4 Intersecting multivariate logistic regression exploring how control over one’s life affects the ability for safer sex based on the intersections between gender and sexual identity

Univariate model

Ability to suggest safer sex

OR b

95% CI c

p-value

Control over one’s life a

No, I had no ability to suggest safer sex at last sex (n = 2244)

1

ref.

ref.

Yes, I had ability to suggest safer sex at last sex (n = 4865)

1.584825

1.263933–1.987185

0.000

Gender * heterosexual identity

Male and heterosexual (n = 2406)

1

ref.

ref.

Female and heterosexual (n = 3890)

1.61278

1.276727–2.037288

0.000

Non-binary and heterosexual (n = 19)

0.599831

0.2120385–2.298937

0.555

Gender * homosexual identity

Male and homosexual (n = 54)

1.891751

0.5623386–6.364

0.303

Female and homosexual (n = 83)

0.1930998

0.1080428–0.345118

0.000

Non-binary and homosexual (n = 3)

Gender * bisexual identity

Male and bisexual (n = 81)

1.700775

0.6344579–4.559221

0.291

Female and bisexual (n = 334)

0.6083382

0.3952046–0.9364146

0.024

Non-binary and bisexual (n = 18)

0.4529464

0.1453062–1.411919

0.172

Gender * open sexual identity

Male and I don’t usually categorize myself sexually (n = 107)

0.581003

0.2519825–1.339635

0.203

Female and I don’t usually categorize myself sexually (n = 239)

0.774941

0.4494872 - 1.336042

0.359

Non-binary and I don’t usually categorize myself sexually (n = 14)

0.5292683

0.0856194 - 3.271748

0.494

1. Exposure:
2. a Control over one’s life: Self-reported variable. The three response alternatives were grouped into two: 1) “Yes”, and 0) “No” or “I am not sure”
3. Statistics:
4. b Likelihood: OR: odds ratio (model 4), CI: 95% confidence interval
5. c Proportions: % are weighted proportions according to UngKAB15 to ensure that the sample group responses are representative of the total population aged 16–29 in Sweden