There are a multitude of myths connected with vapour: how it should be inhaled, what vapour is comprised of, and so on. Some of them are below. This article is probably best read along with the one at:
Inhalation technique with an e-cigarette


1. "No one uses a direct lung inhale with a cigarette"; or, "No one uses a mouth inhale with a cigarette".

Wrong: people do both. Read the thread linked here:
Lung Inhale vs Mouth Inhale

2. "People mainly use a mouth inhale with ecigs", or "People mainly use a direct lung inhale with ecigs".

Wrong. People do both. Who does what may relate to whether or not they are a beginner, or using a mini ecig, or if they did one or the other when smoking.

Beginners should always start with a mouth inhale when switching to ecigs, as there is far less chance of lung irritation. A variety of other methods are used by experienced vapers according to the hardware or their preference.

3. "Ecigs irritate the lungs."

Wrong. If a beginner uses a direct lung inhale, then that is an almost certain recipe for lung irritation. Beginners should use the mouth-first-lung-second inhalation method only. Experts can do as they like, because they will know how to avoid lung irritation.

4. "Ecig vapour is mostly PG/VG and there isn't much water, because there is no water in e-liquid."

Wrong: read the vapour analysis here:
Analysis of electronic cigarette vapor
- item #5

Some glycerine-base refills contain over 10% distilled water in the finished e-liquid (up to 20% is added to a glycerine base to reduce its viscosity and make it suitable for vaping).

The average amount of water found in vapour tests of mini ecigs is 66%, and the PG/VG total content is around 5 or 6%. This is explained by the fact that normal exhalation contains a significant quantity of water vapour (therefore - rather obviously - there must be some in exhaled ecig vapour; so a test that shows none at all is clearly faulty). We don't know anything about vapour produced by more efficient hardware, since at Q1 2014 no one has published an analysis of it.


Competence levels
We need to be very careful about interpreting vapour analyses, as there are dozens of incompetent or deliberately misleading analyses available. The most common errors are:

a. An 'analysis' that does not total 100% or even anywhere near it - for example 70% of the total.

There are a surprising number like this, and it is hard to see what use they are supposed to be.

b. An analysis that presents a list of ingredients totalling exactly 100%.

The problem with this is that a 100% analysis of anything does not exist in commercial situations, it is just too expensive to reach that level of accuracy. A good-quality analysis will always state something like, "...and there were 0.4% unidentifiable compounds". (They provided a 99.6% complete analysis along with the explanation that - at the cost/time level paid for - 0.4% could not be identified.) The best commercial standard is around 99.5% because any lower indicates either a lack of competence or strict cost control, and any higher is not possible commercially as the cost rises exponentially after this point. Most labs offer a c. 98% analysis service, and the cost rises significantly above this.

A '100%' analysis is almost guaranteed to be faulty in some way.

c. An analysis that presents a list of ingredients totalling around 100% but with no mention of water.

Since water must be present (along with anything else normally present in exhaled breath such as CO2, and occasionally present such as traces of formaldehyde), and can comprise up to two-thirds of ecig vapour, any such analysis is fatally flawed. Typically, such an analysis will report a nicotine content of over 5%, which is highly anomalous. When this is adjusted downward for the typical 66% water content, the nicotine content then reads 1% or so, which is correct. Such analyses are common, and represent complete incompetence. (Unless of course a statement is made that the analysis represents only the particulate matter and excludes water; which would mean the percentages are irrelevant in the context of 'what ecig vapour contains'.)

d. An analysis of vapour that includes compounds such as phenols, pthalates, or other compounds anomalous in vapour.

This indicates a problem with lab testing protocols. We have so many tests now that such results can be identified as faulty, and we can see why:
i. Phenol inclusions indicate that the ecig was operated inverted and burnt out, or run dry and burnt out (phenols are a component found in the smoke produced by melted plastic).
ii. Pthalate inclusions indicate that low-grade plastic containers were used in the lab for cartomiser liquid extraction processes such as solvent washing, or solvent washing of vapour extraction vessels, which leached pthalates from the cheap, disposable lab containers used for economy. Glass must be used for all containers used in lab testing where solvents may be present.
Pthalates are plasticisers that are easily leached out of non food grade plastic containers. This is why all e-liquid is sold in food grade plastic bottles, or better yet, glass.
iii. Formaldehyde measurements with in vivo exhaled vapour measurements must be compared with the same measurement for the same person in the same chamber before vapour measurement; one must be subtracted from the other, since people can exhale formaldehyde.


Accuracy of analyses
So we can see that all manner of incompetent labs are permitted to publish analyses in journals; accepting a published vapour analysis at face value is unwise. You will need to see photographs of the testing equipment, photos of the testing procedure, a description of the protocols especially those used for puff extraction and vapour collection, a description of the methodology, explanation of solvent usage, materials utilised for storage and testing, and before/after comparisons for certain types of tests (e.g. for formaldehyde). One thing we do know from published vapour tests is that there is a huge variation in the quality of laboratories and staff; in some cases we can also see that funding sources and agendas have affected the outcome.

An acceptable analysis of vapour will present around a 99.5% ingredient list, and the largest component identified will be water. Any anomalies must be examined along with an explanation of why the protocols or equipment utilised could not have contributed to the abnormal result in any way. There are obvious errors in so many published tests that it is hard to accept any analysis without a detailed analysis of the tests themselves.