Vitamin supplements may increase risk of death

Vitamin supplements taken by millions of people do not increase life expectancy and may raise the risk of a premature death , according to a review of 67 studies with more than 230,000 subjects.

The review, by the Cochrane Collaboration which regularly pools data from trials to evaluate drugs and treatments, found supplements vitamin A, vitamin E and beta-carotene are detrimental to health. In 47 trials with 180,938 people and a low risk of bias, the "antioxidant supplements significantly increased mortality", the authors wrote. When the antioxidants were assessed separately and low risk of bias trials were included and selenium excluded, vitamin A was linked to a 16% increased risk of dying, beta-carotene to a 7% increased risk and vitamin E to a 4% increased risk.

Evidence for vitamin C and selenium was more equivocal, suggesting there was no benefit to taking these pills compared with a placebo.

"The bottom line is current evidence does not support the use of antioxidant supplements in the general healthy population or in patients with certain diseases," said Goran Bjelakovic, who performed the review at Copenhagen Universityhospital in Denmark. "There was no indication that vitamin C and selenium may have positive or negative effects. So regarding these we need more data from randomised trials."

All the supplements are categorised as antioxidants; research has suggested these chemicals underlie some of the beneficial effects of eating fruit and vegetables because they soak up harmful byproducts of metabolism which can damage cells and cause aging.

While the evidence of a beneficial effect of a diet rich in fruit and veg is solid, the Cochrane data suggest antioxidant supplements are either useless or detrimental.

Bjelakovic's team evaluated 67 randomised clinical trials with 232,550 subjects; 21 of the trials were on healthy subjects, while the rest tested patients with a range of diseases. The evidence suggests it would be safer to obtain the chemicals not as supplements but by eating plenty of fruit and vegetables.

Original here

Why There Aren't More Googles

Umair Haque wrote recently that the reason there aren't more Googles is that most startups get bought before they can change the world.

Google, despite serious interest from Microsoft and Yahoo—what must have seemed like lucrative interest at the time—didn't sell out. Google might simply have been nothing but Yahoo's or MSN's search box.

Why isn't it? Because Google had a deeply felt sense of purpose: a conviction to change the world for the better.

This has a nice sound to it, but it isn't true. Google's founders were willing to sell early on. They just wanted more than acquirers were willing to pay.

It was the same with Facebook. They would have sold, but Yahoo blew it by offering too little.

Tip for acquirers: when a startup turns you down, consider raising your offer, because there's a good chance the outrageous price they want will later seem a bargain. [1]

From the evidence I've seen so far, startups that turn down acquisition offers usually end up doing better. Not always, but usually there's a bigger offer coming, or perhaps even an IPO.

Of course, the reason startups do better when they turn down acquisition offers is not necessarily that all such offers undervalue startups. More likely the reason is that the kind of founders who have the balls to turn down a big offer also tend to be very successful. That spirit is exactly what you want in a startup.

While I'm sure Larry and Sergey do want to change the world, at least now, the reason Google survived to become a big, independent company is the same reason Facebook has so far remained independent: acquirers underestimated them.

Corporate M&A is a strange business in that respect. They consistently lose the best deals, because turning down reasonable offers is the most reliable test you could invent for whether a startup will make it big.

VCs

So what's the real reason there aren't more Googles? Curiously enough, it's the same reason Google and Facebook have remained independent: money guys undervalue the most innovative startups.

The reason there aren't more Googles is not that investors encourage innovative startups to sell out, but that they won't even fund them. I've learned a lot about VCs during the 3 years we've been doing Y Combinator, because we often have to work quite closely with them. The most surprising thing I've learned is how conservative they are. VC firms present an image of boldly encouraging innovation. Only a handful actually do, and even they are more conservative in reality than you'd guess from reading their sites.

I used to think of VCs as piratical: bold but unscrupulous. On closer acquaintance they turn out to be more like bureaucrats. They're more upstanding than I used to think (the good ones, at least), but less bold. Maybe the VC industry has changed. Maybe they used to be bolder. But I suspect it's the startup world that has changed, not them. The low cost of starting a startup means the average good bet is a riskier one, but most existing VC firms still operate as if they were investing in hardware startups in 1985.

Howard Aiken said "Don't worry about people stealing your ideas. If your ideas are any good, you'll have to ram them down people's throats." I have a similar feeling when I'm trying to convince VCs to invest in startups Y Combinator has funded. They're terrified of really novel ideas, unless the founders are good enough salesmen to compensate.

But it's the bold ideas that generate the biggest returns. Any really good new idea will seem bad to most people; otherwise someone would already be doing it. And yet most VCs are driven by consensus, not just within their firms, but within the VC community. The biggest factor determining how a VC will feel about your startup is how other VCs feel about it. I doubt they realize it, but this algorithm guarantees they'll miss all the very best ideas. The more people who have to like a new idea, the more outliers you lose.

Whoever the next Google is, they're probably being told right now by VCs to come back when they have more "traction."

Why are VCs so conservative? It's probably a combination of factors. The large size of their investments makes them conservative. Plus they're investing other people's money, which makes them worry they'll get in trouble if they do something risky and it fails. Plus most of them are money guys rather than technical guys, so they don't understand what the startups they're investing in do.

What's Next

The exciting thing about market economies is that stupidity equals opportunity. And so it is in this case. There is a huge, unexploited opportunity in startup investing. Y Combinator funds startups at the very beginning. VCs will fund them once they're already starting to succeed. But between the two there is a substantial gap.

There are companies that will give $20k to a startup that has nothing more than the founders, and there are companies that will give $2 million to a startup that's already taking off, but there aren't enough investors who will give $200k to a startup that seems very promising but still has some things to figure out. This territory is occupied mostly by individual angel investors—people like Andy Bechtolsheim, who gave Google $100k when they seemed promising but still had some things to figure out. I like angels, but there just aren't enough of them, and investing is for most of them a part time job.

And yet as it gets cheaper to start startups, this sparsely occupied territory is becoming more and more valuable. Nowadays a lot of startups don't want to raise multi-million dollar series A rounds. They don't need that much money, and they don't want the hassles that come with it. The median startup coming out of Y Combinator wants to raise $250-500k. When they go to VC firms they have to ask for more because they know VCs aren't interested in such small deals.

VCs are money managers. They're looking for ways to put large sums to work. But the startup world is evolving away from their current model.

Startups have gotten cheaper. That means they want less money, but also that there are more of them. So you can still get large returns on large amounts of money; you just have to spread it more broadly.

I've tried to explain this to VC firms. Instead of making one $2 million investment, make five $400k investments. Would that mean sitting on too many boards? Don't sit on their boards. Would that mean too much due diligence? Do less. If you're investing at a tenth the valuation, you only have to be a tenth as sure.

It seems obvious. But I've proposed to several VC firms that they set aside some money and designate one partner to make more, smaller bets, and they react as if I'd proposed the partners all get nose rings. It's remarkable how wedded they are to their standard m.o.

But there is a big opportunity here, and one way or the other it's going to get filled. Either VCs will evolve down into this gap or, more likely, new investors will appear to fill it. That will be a good thing when it happens, because these new investors will be compelled by the structure of the investments they make to be ten times bolder than present day VCs. And that will get us a lot more Googles. At least, as long as acquirers remain stupid.

Original here

Road Safety: The Uncrashable Car?

Look in your blind spot, too! The 'uncrashable car' is part of basic research undertaken by the largest research initiative into road safety ever undertaken in Europe. (Credit: PReVENT, Volvo)

The largest road safety research project ever launched in Europe will usher in a series of powerful road-safety systems for European cars. But, in the long term, its basic, experimental research could lead to a car that is virtually uncrashable.

A truck exits suddenly from a side road, directly into your lane only dozens of metres ahead. Suddenly, your car issues a warning, starts applying the brakes and attempts to take evasive action. Realising impact is unavoidable; in-car safety systems pre-tension the safety belts and arm the airbag, timing its release to the second before impact.

Such is the promise of the uncrashable car, coming to a dealer near you in the perhaps not-too-distant future. The system is part of the basic research undertaken by the largest research initiative into road safety ever undertaken in Europe.

PReVENT has a budget of over €50 million and 56 partners pursuing a broad, but highly complementary programme of research. A dozen sub-projects focus on specific road-safety issues, but all projects support and feed into each other in some way.

PReVENT is studying relatively cheap, even simple, technologies – such as parking sensors and existing satellite navigation – that can be retooled to enhance driver safety. But as part of its broad and deep approach to car safety, it is also diving into more experimental and medium- to long-term systems, innovations that could appear in five-to-ten years.

The uncrashable car is a theoretical construct that concerned a handful of PReVENT’s sub-projects. But it could become far more of a reality than anyone expected.

Of course, it is impossible to stop all car collisions, but the technology could be pushed to make it increasingly unlikely and mitigate crashes when they do occur.

For example, PReVENT project WILLWARN uses wireless communication with other vehicles to alert the driver about potentially dangerous situations ahead, while MAPS&ADAS reads sat-nav maps to track approaching hazards, like bends, dips or intersections. SASPENCE looks at safe driving distances and speed, while LATERALSAFE finally brings active sensing to the blind spot.

All have their role in the uncrashable car, as do many others within the broader project. But two projects, APALACI and COMPOSE, take this a step further, actively tracking the speed and trajectories of surrounding vehicles and other road users in real time. If one vehicle suddenly stops, or a pedestrian suddenly steps onto the road, they swing into action to rapidly calculate the implications.

Predictive collision detection

APALACI is an advanced pre-crash mitigation system built round the registration of other motorists and cyclists. In the APALACI system, sensors monitor the street or road immediately around the vehicle and collect as much information about a collision as possible, before it even starts to take place.

The system uses this data to decide on the ideal safety reaction strategy. Examples include controlled braking manoeuvres, controlled activation of the occupant restraint systems or pre-arming airbag systems. The car can react far faster than the driver, cutting speed by crucial amounts to ensure unavoidable accidents are less severe.

APALACI also developed a so-called ‘Start Inhibit System’ for trucks. It surveys the blind spot immediately in front of a truck and protects pedestrians or cyclists by preventing dangerous manoeuvres.

APALACI was tested in a series of vehicles like the Fiat Stilo, the Volvo FH12 truck, the Alfa Romeo 156 and Mercedes E350. It used laser sensors, radar, software decision assistance and a variety of other technologies to achieve the goal.

Tiny changes have a huge impact

COMPOSE, on the other hand, aims more specifically to keep others, as well as its driver, safe. It can apply the brakes if a pedestrian steps onto the road, or extend the bumper, and raise the bonnet to enhance occupant protection.

Tiny differences have a huge impact on car safety. Dropping speed by 1km/h can reduce accidents with injury by 3 per cent, while braking fractions of a second sooner is enough to reduce the damage caused dramatically.

The systems were tested in the BMW 545i and the Volvo FH12 truck, and they do appreciably enhance safety. But, for all their potential, these systems remain, for now, the preserve of the future.

“The teams developed sophisticated algorithms to track all these elements in the landscape,” explains Matthias Schulze, coordinator of the EU-funded PReVENT project and Senior Manager for ITS & Services at Daimler AG. “But they require enormous computer power to keep track of all the various elements, so this work is aimed at basic research, establishing how it could be done. It will be a while before in-car computers are sophisticated enough to use these systems.”

Nonetheless, they do provide tools that automakers can use to mitigate the potential for accidents, and they provide a clear research roadmap for the uncrashable car of the future.

Original here