A Czech study involving 146 people with mild to moderate risk of cardiovascular disease has found that moderate wine drinking was only protective in people who exercised.
Participants were randomly assigned to one year of moderate consumption of red wine (Pinot Noir) or white wine (Chardonnay-Pinot) from the same year and wine region of the Czech Republic. They had to keep a logbook on their consumption of wine and other alcoholic beverages, medication use, and amount and type of exercise. They were required to return the corks from the wine bottles to confirm that they had drank the wine rather than sold it.
The study was looking for a rise in HDL cholesterol, indicating a protective effect against cardiovascular disease. However, there was no change in HDL cholesterol levels by the end of the study in either the red or white wine groups. Nor did it appreciably affect blood glucose, triglycerides, or levels of inflammatory markers like C-reactive protein, although LDL cholesterol was lower in both groups while total cholesterol was lower only in the red wine group.
However, those participants who engaged in regular exercise at least twice a week did show an increase in HDL cholesterol, as well as a decrease in LDL and total cholesterol. This occurred in both red and white wine groups.
Of course, the study didn't include a control group of people engaging in regular exercise who didn't drink, so all we can really say (and it's certainly worth saying) is that moderate wine drinking doesn't seem to help protect against cardiovascular disease, and that despite all the hype around red wine, with its greater antioxidants and resveratol, there's no evidence it's any healthier than white wine.
http://www.eurekalert.org/pub_releases/2014-08/esoc-wop082614.php
http://www.theatlantic.com/health/archive/2014/09/working-with-the-wine-not-against-it/379504/
Taborsky M, Ostadal P, Petrek M. A pilot randomized trial comparing long-term effects of red and white wines on biomarkers of atherosclerosis (in vino veritas: IVV trial). Bratisl Lek Listy. 2012;113(3):156-158.