The argument that veg*s enjoy better health because of the confounding factors (reduced smoking habit, reduced alcohol consumption, more physical activity, etc.) could be valid for cohort studies, but there are actually several clinical trials (RCT and even cross-over) that show that - although there are some health benefits derived from these behaviours - a part of the health improvements are due to the diet alone.
Here are some examples:
In Barnard 2009 (RCT), 99 patients with type 2 diabetes were
randomly assigned to a low-fat vegan diet (n = 49) or a diet following
2003 American Diabetes Association guidelines (conventional, n = 50)
for 74 weeks. The low-fat vegan diet appeared to improve glycemia and
plasma lipids more than did conventional diabetes diet
In Barnard 2005 (RCT), 64 overweight, postmenopausal women were
randomly assigned to a low-fat, vegan diet or a control diet based on
National Cholesterol Education Program guidelines, without energy
intake limits, and were asked to maintain exercise unchanged. The
group on the low-fat vegan diet improved their body weight,
insulinemia and glicemia more than the other.
In Kahleova 2011 (RCT), 74 patients with Type 2 diabetes were
randomly assigned to either the experimental group (n = 37), which
received a vegetarian diet, or the control group (n = 37), which
received a conventional diabetic diet. Several improvements where
higher in the first group than the second: BMI reduction, waist
circumference, body weight, total and LDL cholesterol, homocisteine,
glicemia, insulinemia, HbA1c.
In Siener 2003 (cross-over), 10 healthy male subjects ingested a
self-selected meat-containing diet (SD) for two weeks, and three
different standardized diets for a period of 5 days each. The
Westerntype diet (WD) was representative of the usual dietary
habits,whereas the balanced omnivorous diet (OD) and the
ovo-lacto-vegetarian diet (VD) were calculated according to the
requirements. The risk of uric acid crystallization was highest on the
ingestion of diets SD and WD, due to the high urinary uric acid
excretion and the acidic urinary pH. The relative supersaturation with
uric acid declined significantly by 85% on the intake of diet OD,
consequent to the decrease in uric acid excretion and concentration
and the increase in urinary pH value. The ingestion of the vegetarian
diet VD led to a further significant reduction in the risk of uric
acid crystallization by 93% compared to diet WD.
In Cooper 1982 (cross-over), 15 free-living individuals were
randomly assigned to 3-week periods on either the experimental diet or
a control diet which closely approximated the usual intake in the
U.S.A. Significant reductions in total cholesterol (12.5%), low
density lipoprotein cholesterol (14.7%), and apo B (13.2%) were
In Bunner 2014 (cross-over), 42 adult migraine sufferers were
divided randomly into two groups. This 36-week crossover study
included two treatments: dietary instruction and placebo supplement.
Each treatment period was 16 weeks, with a 4-week washout between.
During the diet period, a low-fat vegan diet was prescribed for 4
weeks, after which an elimination diet was used. The patients showed
greater improvement during the diet period than placebo.