This study seems pretty legit to me.
http://ntr.oxfordjournals.org/content/early/2014/05/14/ntr.ntu078.full
I copy and paste some of it here.
"> nicotine & tobacco research, volume 16, number 10 (October 2014)
1319–
1326
Carbonyl Compounds in electronic Cigarette vapors:
effects of nicotine solvent and Battery Output voltage
Leon Kosmider PharmD1,2, Andrzej Sobczak PhD1,2, Maciej Fik PharmD2, Jakub Knysak PharmD2,
Marzena Zaciera PhD1, Jolanta Kurek PhD1, Maciej Lukasz Goniewicz PharmD, PhD3
http://ntr.oxfordjournals.org/
Electronic Cigarette
The most popular device available on the Polish market as on
January 2013 was selected for the study. Because the Internet
is currently the main distribution channel for EC, we searched
google.pl web browser and tracked the number of EC sell offers
on Allegro.pl, which is the most popular online auction service
in Poland. Based on the number of search hits and sell offers,
we chose and purchased the eGo-3 brand (Volish, Ltd, Poland).
The device has controlled maximum time for single puff of
10 s. We chose a model composed of a Crystal 2 clearomizer
with a heating element with resistance of 2.4 ohms, a 900 mAh battery
with voltage of 3.4V, and a battery voltage stabilization system.
Results
Levels of Carbonyl Compounds Released From
Commercially Available Refill Solutions
Table 2 shows amounts of each analyzed carbonyl compounds
in 15 puffs of vapor from 10 commercially available e-liquids.
The values presented in Table 2 are means with
SD from three
tests performed at the same voltage of 3.4V. All samples contained
at least one carbonyl compound. Formaldehyde, acetaldehyde,
acetone, and butanal were found in most of the
analyzed samples. However, not all commercially available
e-liquids emitted all these four carbonyls. Crotonaldehyde was
detected in only one sample (A10), whereas acrolein was not
detected in any sample.
The amounts of formaldehyde and acetaldehyde in vapors from lower voltage EC were on average 13- and 807-fold lower than in tobacco smoke, respectively.
"> The striking finding of our study is that levels of carbonyls rapidly increase with increased battery output voltage.
Increasing battery output voltage leads to higher temperature of the heating element inside EC. In addition, the increased
battery output voltage results in more e-liquid consumed per puff. Our findings show that increasing voltage from 3.2 to
4.8V resulted in 4 to over 200 times increase in formaldehyde, acetaldehyde, and acetone levels. The levels of formaldehyde in vapors from high-voltage devices were in the range of levels reported in tobacco smoke (1.6–52 µg/cigarette;
Counts, Morton, Laffoon, Cox, & Lipowicz, 2005).
This finding suggests that in certain conditions ECs might expose their
users to the same or even higher levels of carcinogenic formaldehyde than tobacco smoke. This finding is essential for
the product safety and in the light of forthcoming regulation of the devices.
The amounts of formaldehyde and acetaldehyde in vapors from lower voltage EC were on average 13- and 807-fold lower than in tobacco smoke, respectively.
Conclusions
Vapors from ECs contain toxic and carcinogenic carbonyl compounds. Both solvent and battery output voltage significantly
affect levels of carbonyl compounds in EC vapors. Levels of carbonyls rapidly increase with increased battery output voltage. New generation of high-voltage ECs may put their users in increased health risk from exposure to high levels of carbonyl compounds although the risk
will still probably be much lower compared with smoking.