Lastly, don`t kid yourself about helmets. Too many people think of these things as some sort of magic talisman that makes you invincible and as a bonus wards off evil. Most helmets are rated for under 15 MPH and an impact more than that makes the helmet increasingly ineffective. If you are riding at 25 MPH, which is really slow for any intermediate rider, and you slam your head to the ice or hit a tree, you are getting a concussion helmet or not.
If you hit a solid object with your head at 35mph, you are likely dead - helmet or not. But, if you crash on an icy slope, the helmet doesn't have to absorb your movement speed, it primarily has to absorb the fall component. I.e. if your head falls 2m (you topple over), it accelerates to 14mph.
The concussion is caused by the sudden stop and the brain slamming into the skull. This still happens with a helmet.
Not really. The helmet is designed to cushion the 'sudden stop' - make it not so sudden - and absorb impact energy.
Excerpt from an old Ski Canada Magazine article, "The Science Behind Helmets":
An insight as to why this study found a difference in patterns of death as a function of helmet utilization can be found in the following study. A simulation using a 50th percentile male anthropometric device (Scher, Richards and Carhart, 2005) was done of a snowboarder going 30 kph, catching an edge and falling headfirst onto soft snow, icy snow and a fixed object (a 28-cm upright wooden post). This simulation was done to assess the effect of wearing a helmet or not under the three different impact conditions. The helmet in question met the requirements of ASTM F2040. [...] This study found that if the impact is onto a soft-snow surface, both the measured g-loads (under 100 g) and the computed HIC values (less than 220) are well within acceptable limits regardless of whether or not a helmet is used. When the impact was onto simulated hard, icy snow, the helmet reduced the average measured g-load from 329 to 162, and the HIC value from 2,235 to 965. When the impact was against the fixed object, the helmet reduced the values from 696 to 333, and the HIC from 12,185 to 3,299.
The study concluded that under the circumstances of impact with soft snow, the use or non-use of the helmet had no significant effect. In the matter of the impact with a solid fixed object resembling a tree, while the use of a helmet was associated with a significant reduction in both the g-load and the HIC, the likely outcome remained that of a fatal injury— with or without the use of a helmet. With an impact on icy snow, the use of a helmet could be the difference between a significant head injury (possibly life-threatening) and a minor head injury.
There is a huge difference between helmets, even if they are certified. The standards allow something like 250g - 300g deceleration. Above 100g, chances are you will get a concussion. Above 300g, chances are you will have fractured head.
There is a large number number of studies that show that helmets do reduce head injuries for snowboarders. A lot of people seem to think they will perform miracles - which they clearly don't.
While I do wear a helmet for boarding (I've hit my head a few times in soft snow, fortunately never on hard pack - knock on wood), I don't use a bicycle helmet. IMHO, there isn't enough evidence that bicycle helmets do something useful.