E-cigarette liquids: Constancy of content across batches and accuracy of labeling

•E-liquids are used by millions of consumers, but few controls were in place until recently.•This study shows that an online sample of bottles of refill liquids were filled true to label.•The content of these e-liquid bottles was stable across two production batches.•The samples did not contain impurities.•Tests are needed for a longer list of substances, both in the liquids and in the aerosols.

AimsTo assess whether bottles of refill liquids for e-cigarettes were filled true to label, whether their content was constant across two production batches, and whether they contained impurities.MethodsIn 2013, we purchased on the Internet 18 models from 11 brands of e-liquids. We purchased a second sample of the same models 4 months later. We analyzed their content in nicotine, anabasine, propylene glycol, glycerol, ethylene glycol and diethylene glycol, and tested their pH.ResultsThe median difference between the nicotine value on the labels and the nicotine content in the bottles was 0.3 mg/mL (range − 5.4 to + 3.5 mg/mL, i.e. − 8% to + 30%). For 82% of the samples, the actual nicotine content was within 10% of the value on the labels. All models contained glycerol (median 407 mg/mL), and all but three models contained propylene glycol (median 650 mg/mL). For all samples, levels of anabasine, ethylene glycol and diethylene glycol were below our limits of detection. The pH of all the e-liquids was alkaline (median pH = 9.1; range 8.1 to 9.9). The measured content of two batches of the same model varied by a median of 0% across batches for propylene glycol, 1% for glycerol, 0% for pH, and 0.5% for nicotine (range − 15% to + 21%; 5th and 95th percentiles: − 15% and + 10%).ConclusionsThe nicotine content of these e-liquids matched the labels on the bottles, and was relatively constant across production batches. The content of propylene glycol and glycerol was also stable across batches, as was the pH.