Objective The WHO Framework Convention on Tobacco Control and its Guidelines recommend nations ban the use of misleading terms, such as ’light' and ’mild' on tobacco product advertising, packaging and labelling. Many nations, including the USA, have implemented such bans and some have introduced or passed legislation requiring plain packaging on tobacco products. We previously reported that manufacturers in the USA responded by replacing lights terms with colour terms and related colour-coding of packages. This study examines population outcomes and public health impact of the US ban.

Methods We examined available data regarding a) per cent filter ventilation strata used to designate lights subbrand categories; b) market share per tar yield; c) initiation and use of cigarettes by lights categories and d) overall cigarette consumption to identify changes from before to after the ban. We used interrupted time series multivariable logistic regression and joinpoint regression models to test for changes in rates and temporal trends associated with the ban.

Results The per cent filter ventilation strata used to designate lights subbrand categories were maintained in the colour named subbrands. No change was observed following the ban in lights market share, relative prevalence of lights versus non-lights smoking or relative smoking initiation on lights versus non-lights among all ages or among youth in particular. The rate of decline in per capita cigarette consumption slowed by 37% from the period 2007-2010 to 2010-2014.

Conclusions This study strongly suggests that manufacturers' circumvention prevented the lights descriptor ban from succeeding as intended, most likely perpetuating the misleading consumer perceptions about relative risks, while failing to increase smoking cessation and reduce initiation. Laws requiring generic (plain) and elimination of subbrand descriptors should prevent evasion of legislation banning the use of specific terms through marketing, regulatory and legal challenges.