“Research highlights” is a weekly round-up of research papers appearing in the print BMJ. We start off with this week’s research questions, before providing more detail on some individual research papers and accompanying articles.

• How did breast cancer mortality in the United Kingdom and other parts of Europe change between 1989 and 2006?
• Which risk factors should be avoided to prevent dementia?
• How long do people survive after diagnosis of dementia in primary care?
• Does ambient outdoor temperature affect the short term risk of myocardial infarction?

Dementia risks and survival

Karen Ritchie and colleagues reviewed the evidence on risk factors for dementia and then modelled, using data from a prospective cohort study of older people in Montpelier, France, the impact on dementia incidence if each of these risks was addressed completely. Their final model estimated that, for people in their early seventies, increasing literacy might cut dementia incidence in the next seven years by 18% (rather than “literacy” the authors use the intriguing term “crystallised intelligence”, which sounds like something from science fiction but was assessed using the Neale adult reading test). Furthermore, eliminating depression might reduce the risk by 10%, increasing fruit and vegetable consumption by 6.5%, and eliminating diabetes by 4.8%. Stroke, heart disease, head injury, and other factors were included in the analyses at first but didn’t make it into the final model.

This was an interesting and rigorous study, but was entirely hypothetical. The big challenge, as editorialists Tobias Kurth and Giancarlo Logroscino point out, is to translate the findings into prevention. There’s still a long way to go before that will be possible.

Two years ago we published Jing Xie and colleagues’ population based cohort study from England and Wales showing that people with dementia, newly diagnosed by researchers using a formal algorithm, went on to survive 4.5 years on average (doi:10.1136/bmj.39433.616678.25).

Unsurprisingly, survival varied with age, from about 10 years for people in their late 60s to 3.8 years for the over 90s. But most people with dementia, at least in the UK, are diagnosed and mostly managed in primary care. Greta Rait and colleagues’ cohort study using a large general practice database found median survival rates of 6.7 years for 60-69 year olds and less than two years for the over 90s. Dementia may well be under-recognised in primary care, or at least under-reported. Either way, this new evidence should help general practitioners to discuss prognosis with patients and their carers.

Disparities in breast cancer mortality in Europe

In international surveys of cancer survival, the UK often seems to compare poorly with other European countries (BMJ 2007;335:363). But Phillipe Autier and colleagues have some better news. Their analyses of data from the World Health Organization mortality database show that breast cancer mortality fell by more than 20% between 1989 and 2006 in two thirds of European countries, and that England and Wales were beaten only by Iceland in the league table of countries with falling rates. Rates in France and Sweden changed little, however, despite extra investments in screening and treatment.

Editorialists Valerie Beral and Richard Peto discuss why survival estimates based on UK registry data can make the lot of a cancer patient here look worse than it really is.

Under the weather: falling outdoor temperatures and myocardial infarction

We know that heat waves and cold snaps can be lethal for frail older people. But how does the risk of myocardial infarction vary with ambient outdoor temperature? To answer this, Krishnan Bhaskaran and colleagues analysed data from a national audit of hospital admissions for myocardial infarction in 15 urban areas in England and Wales and from daily weather reports. They came to the neat and memorable conclusion that each 1°C drop in daily mean temperature was associated with a 2% increased risk of myocardial infarction. But don’t try to extrapolate this to countries with different climates, warn the authors.

Research online: For these and other new research articles see www.bmj.com/channels/research.dtl

Concerns over calcium supplementation

Last week we published a meta-analysis that linked calcium supplementation to an increased risk of myocardial infarction in middle aged people (BMJ 2010;341:c3691), along with an editorial (BMJ 2010;341:c3856). Despite the authors’ cautious discussion of the limitations, rapid responders are unhappy about the studies included in the meta-analysis (http://www.bmj.com/cgi/eletters/341/jul29_1/c3691). Ediriweera Desapriya and Namrata Dutt from Vancouver, Canada, write: “We do not know that 11 studies in this meta-analysis were generally of good quality and had no selection or attribution biases,” whereas Soaring Bear says: “I would also have to note that 30% in a meta-analysis of heterogeneous studies is not very much and raises the question of significance.” Consultant rheumatologist Matthew Grove bemoans the exclusion of studies with concomitant vitamin D. “Only 3.5% of our prescriptions [for 2009-10] were for calcium supplements not containing vitamin D,” he says.