Table 3

Logistic regressions of message condition on perceived comparative harm, comparative addictiveness of VLNCs, and support for VLNC policy

Outcomes/conditionsVLNC message, N=468E-cigarette message, n=484Combined message, n=476Control, n=473
Perceived VLNCs as less harmful than regular cigarettes*
OR0.40 (0.29–0.57)0.85 (0.63–1.14)0.68 (0.50–0.93)Reference
% of ‘less harmful’13.323.320.827.1
% of ‘don’t know’15.717.113.214.4
Perceived VLNCs as less addictive than regular cigarettes*
OR1.05 (0.79–1.40)0.84 (0.63–1.12)1.01 (0.76–1.35)Reference
% of ‘less addictive’32.129.032.733.0
% of ‘don’t know’20.117.217.115.6
Support for VLNC policy†
OR1.09 (0.83–1.42)0.98 (0.75–1.29)1.14 (0.87–1.48)Reference
% of ‘support’39.135.539.337.7
% of ‘don’t know’4.57.66.34.7
Session 2 Outcome/conditionsVLNC message, N=379E-cigarette message, n=383Combined message, n=374Control, n=374
Perceived VLNCs as less harmful than regular cigarettes at Session 2*
OR0.81 (0.57–1.14)1.06 (0.76–1.48)0.94 (0.67–1.32)Reference
% of ‘less harmful’23.528.326.327.1
% of ‘don’t know’20.820.520.221.2
  • Bold indicates significance at adjusted p<0.05.

  • The correlations between the outcomes and potential covariates were consistently below 0.30. Therefore, the models were unadjusted for covariates, and only ORs were presented.

  • *Perceived comparative harm/addictiveness of VLNCs was dichotomised as 1=‘equally or more harmful/addictive’ (equally harmful/addictive, more harmful/addictive and much more harmful/addictive) versus 2=‘less harmful/addictive’ (much less harmful/addictive and less harmful/addictive).

  • †Support for VLNC policy was dichotomised as 1=‘support’ (strongly support, support) versus 2 ‘oppose or neutral’ (strongly oppose, oppose, neither support nor oppose).

  • VLNC, very low nicotine cigarettes.