More people are being diagnosed with eating disorders every year and the most common type is not either of the two most well known—bulimia or anorexia—but eating disorders not otherwise specified (eating disorders that don’t quite reach the threshold to be defined as anorexia or bulimia), shows a study published online in BMJ Open.

Few studies have investigated the incidence of eating disorders, so the authors set out to determine the incidence of diagnosed anorexia nervosa, bulimia nervosa and other non-specified eating disorders in primary care over a 10-year period in the UK (2000–2009), to see how the incidence had changed and the most common age of diagnosis.

Eating disorders have the highest death rates of all mental disorders, and understanding changes in their incidence over time and by age and gender is essential to ensure timely diagnosis and appropriate service provision.

The researchers used data from the General Practice Database, which contains anonymised records of 5% of the UK population to identify all cases of eating disorder diagnosed between 2000 and 2009. They found a total of 9072 cases.

Analysis of the data revealed that in 2000 there were 32.3 new cases of eating disorder per 100,000 population aged between 10 and 49 years, and that this rose steadily to 37.2 new cases per 100,000 by 2009.

Despite other research suggesting a decrease in the incidence of bulimia, the incidence of bulimia and anorexia remained stable over the 10 years and it was new cases of eating disorder not otherwise specified which were responsible for the overall rise. The incidence of these unspecified eating disorders, which have been far less studied than bulimia and anorexia and are the most common type seen in hospital care (60% of cases seen in specialist services), has not been estimated previously in general practice care. However, publication of the new Diagnostic and Statistical Manual of Mental Disorders (DSM 5) last week is expected to mean that the majority of these cases will be diagnosed with anorexia, bulimia or a new condition of binge eating disorder in future.

The data showed that girls aged 15 to 19 years and boys aged 10 to 14 had the highest incidences of new diagnoses of eating disorder.

Two girls in every 1000 aged 15 to 19 years are likely to be diagnosed with an eating disorder every year, which means that there are around 4610 new cases in girls of this age group each year. As a result, eating disorder is probably the most common new onset mental health disorder in adolescent girls after depression, the authors say. In 2009, the incidence of new diagnoses of depression was 11.9 per 1000 in girls of this age group.

In girls aged 10-19 years, there are nine times as many new cases of eating disorder (1.2 per 1000 population) diagnosed every year as there are cases of type 1 diabetes (0.26 new cases per 1000), and about half as many new diagnoses as there are of type 2 diabetes (3.6 new cases per 1000).

Bed sharing with parents is linked to a fivefold increased risk of sudden infant death syndrome (SIDS), even when the parents are non-smokers and the mother has not been drinking alcohol and does not use illegal drugs, according to a large analysis published online in BMJ Open.

While the rate of SIDS has fallen sharply following advice to parents to place babies to sleep on their back (supine), SIDS remains the major cause of infant death in the postneonatal period (28 days through to the first birthday) in developed countries.

Some countries, such as the Netherlands and the USA, advise parents not to sleep in the same bed as infants less than three months old, whereas others, such as the UK and Australia, advise only certain parents not to bed share with their young infants, including smokers and those who have been drinking alcohol or taking drugs.

The authors of this analysis estimate that around 88% of all SIDS deaths while bed sharing would not have occurred if bed sharing had been avoided. Their results show that even when neither parent smoked, and the baby was less than 3 months old, breastfed and the mother did not drink or take drugs, the risk of SIDS was five times higher than if the baby had slept in a cot next to their parents’ bed.

The risk of SIDS while bed sharing decreased as the age of the infant rose, but if either parent was a smoker or the mother had drunk alcohol (two or more units in the last 24 hours) or used illegal drugs, including cannabis, at any time since the child was born, the risk was greatly increased.

Risks of bed sharing had been reported in different ways, so Professor Bob Carpenter, of the London School of Hygiene and Tropical Medicine, led this study which combined individual data from five published data sets from the UK, Europe and Australasia. It includes data on 1472 SIDS cases and 4679 controls and is the largest ever individual level study of the problem.

It revealed that one or both parents of 22.2% of the infants who had died from SIDS had been sleeping with their child at the time of death, while 9.6% of the parents in the control group had awoken the morning of the interview in the same bed as their child. Over the past 10 years, there has been a marked increase in bed sharing and the authors now estimate that around 50% of SIDS cases occur while bed sharing, more than double the figure found in the study.

The authors write: “88% of the deaths that occurred while bed sharing would probably not have occurred had the baby been placed on its back in a cot by the parents’ bed.” Even in very low-risk breastfed babies, where there were no risk factors for SIDS other than that they had slept in their parents’ bed, 81% of SIDS deaths in infants under three months of age could have been prevented by not bed sharing, they add.

“The current messages saying that bed sharing is dangerous only if you or your partner are smokers, have been drinking alcohol or taking drugs that make you drowsy, are very tired or the baby is premature or of low-birth weight, are not effective,” they say and call for recommendations “that take a more definitive stance against bed sharing for babies under 3 months”.

“We do not suggest that babies should not be brought into the parent’s bed for comfort and feeding. This has been investigated in previous studies and has not been found to be a risk factor, provided the infant is returned to his or her own cot for sleep,” they write.

The most-read article in April was Yamada and colleagues’ analysis of male pattern baldness and coronary heart disease, which received widespread press coverage. Also popular was Küster et al‘s study highlighting the dangers of consuming over the counter analgesics before running a marathon which was also covered in the press in the run-up to the London marathon this month. Other newly-published papers in the top ten include Lin et al‘s study of lower back pain and Mahtani et al‘s study on sport participation following the Olympics.

Most read figures are based on pdf downloads and full text views. Abstract views are excluded.

Young men who are obese in their early 20s are significantly more likely to develop serious ill health by the time they reach middle age, or not even make it that far, suggests research published in BMJ Open.

It’s well known that obesity in adulthood poses a risk for diabetes and cardiovascular disease, but it’s not been clear whether obesity in early adulthood strengthens that risk.

The authors tracked the health of 6500 Danish 22 year old men for 33 years up to the age of 55. All of them had been born in 1955, and had registered with the Military Board for a fitness test to gauge their suitability for military service.

All potential conscripts in Denmark are subjected to a battery of psychological and physical tests, including weight. Most (83%; 5407) were within the normal range and 5% were underweight (353). One in 10 (639) were overweight and 1.5% (97) were obese.

Normal weight is classified as a body mass index (BMI) of between 18.5 and 25; obesity is classified as a BMI of 30 or more.

Almost half of those classified as obese at the age of 22 were diagnosed with diabetes, high blood pressure, heart attack, stroke, blood clots in the legs or lungs, or had died before reaching the age of 55.

They were eight times as likely to get diabetes as their normal weight peers and four times as likely to get a potentially fatal blood clot (venous thromboembolism). They were also more than twice as likely to develop high blood pressure, have had a heart attack, or to have died.

Every unit increase in BMI corresponded to an increased heart attack rate of 5%, high blood pressure and blood clot rates of 10%, and an increased diabetes rate of 20%.

In all, obese young men were three times as likely to get any of these serious conditions as their normal weight peers by middle age, conferring an absolute risk of almost 50% compared with only 20% among their normal weight peers.

The findings prompt the authors to warn that the continuing rise in obesity may counteract the fall in deaths from heart disease.

“Thus, obesity related morbidity and mortality will, in decades to come, place an unprecedented burden on healthcare systems worldwide,” they suggest.

Attempts to ward off pain in marathons and other endurance sports by taking over the counter painkillers may be ill advised, because these drugs may cause serious side effects in these circumstances, suggests research published in BMJ Open.

Many competitors try to prevent pain interfering with their performance by taking painkillers that are readily available in pharmacies and supermarkets, say the authors.

And in a bid to find out what impact these common drugs might have, the authors quizzed participants in the 2010 Bonn Marathon/Half-Marathon about their use of medication and any symptoms they had during and/or after the race.

In all, just under 4000 (56%) of all 7048 competitors returned their online questionnaires. Most (87%) had run marathons before.

Of those who took painkillers before the race, 1 in 5 said they also used painkillers during training to curb or ward off pain; 1 in 10 said they had pain symptoms before the start of the race. This compares with 1% of those who didn’t touch painkillers.

Over half the drugs taken (54%) were bought over the counter without a prescription, and included diclofenac, asprin, and ibuprofen.

The numbers of those forced to withdraw during the race because of pain and other health symptoms differed little between those who took painkillers and those who didn’t.

But withdrawal from competition because of gastrointestinal problems was significantly more common among those taking painkillers, and while withdrawal as a result of muscle cramps was rare, it was significantly more common in those who had taken these drugs.

Runners who popped painkillers were five times as likely to experience symptoms as those who didn’t take these drugs; the overall difference in risk was 13%.

Symptoms included stomach cramps, cardiovascular problems, gastrointestinal bleeds, blood in the urine and joint and muscle pain.

The rate of symptoms rose in parallel with increasing dose. One in 10 of those taking diclofenac took over 100 mg; 43% of those who took ibuprofen, the second most popular choice, took doses of 800 mg or more—twice the recommended dose.

Virtually none of the respondents said they were aware of any risks associated with taking painkillers for endurance sports.

Nine runners who took painkillers said they had ended up in hospital: three for temporary kidney failure after taking ibuprofen; four for bleeding ulcers (aspirin), and two after a heart attack (aspirin), one of whom had taken 500 mg for mild foot pain.

None of the runners who competed without having topped up on painkillers was admitted to hospital.

Painkillers block enzymes called cyclooxygenases, which regulate the production of prostaglandins. But the authors suggest that prostaglandins also protect tissues when the body is under extreme stress, such as during endurance sports.

“Taken together, our data indicate that the widespread use of cyclooxygenase inhibitors in connection with endurance sports is potentially damaging. Further investigations are warranted to examine whether the use of analgesics before and during sports activities should be avoided altogether,” they conclude.

The most-read article in March was Christine Clar and colleagues’ systematic review of SGLT2 receptor inhibitors in type 2 diabetes. Doyle et al‘s  review on the links between patient experience and clinical safety – originally published in January – remains popular,  and Katzmarzyk et al’s article discusses sedentary behaviour and life expectancy in the USA also proves popular as the third most-read article this month.

Most read figures are based on pdf downloads and full text views. Abstract views are excluded.

Male pattern baldness is linked to an increased risk of coronary heart disease, but only if it’s on the top/crown of the head, rather than at the front, finds an analysis of published evidence in BMJ Open.

A receding hairline is not linked to an increased risk, the analysis indicates.

The researchers trawled the Medline and the Cochrane Library databases for research published on male pattern baldness and coronary heart disease, and came up with 850 possible studies, published between 1950 and 2012.

But only six satisfied all the eligibility criteria and so were included in the analysis. All had been published between 1993 and 2008, and involved just under 40,000 men.

Three of the studies were cohort studies – meaning that the health of balding men was tracked for at least 11 years.

Analysis of the findings from these showed that men who had lost most of their hair were a third more likely (32%) to develop coronary artery disease than their peers who retained a full head of hair.

When the analysis was confined to men under the age of 55-60, a similar pattern emerged. Bald or extensively balding men were 44% more likely to develop coronary artery disease.

Analysis of the other three studies, which compared the heart health of those who were bald / balding with those who were not, painted a similar picture.

It showed that balding men were 70% more likely to have heart disease, and those in younger age groups were 84% more likely to do so.

Three studies assessed the degree of baldness using a validated scale (Hamilton scale). Analysis of these results indicated that the risk of coronary artery disease depended on baldness severity, but only if this was on the top/crown of the head, known as the vertex.

Extensive vertex baldness boosted the risk by 48%, moderate vertex baldness by 36%, and mild vertex baldness by 18%. By contrast, a receding hairline made very little difference to risk, the analysis showed.

To compensate for differences in the methods of assessing baldness in the studies included in the analysis, the authors looked at four differing grades of baldness: none; frontal; crown-top; combined.

Once again, this indicated that the severity of baldness affected the risk of coronary heart disease.

Men with both frontal and crown-top baldness were 69% more likely to have coronary artery disease than those with a full head of hair, while those with just crown-top baldness were 52% more likely to do so. Those with just frontal baldness were 22% more likely to do so.

Explanations for the reasons behind the association vary, but include the possibility that baldness may indicate insulin resistance, a precursor to diabetes; a state of chronic inflammation; or increased sensitivity to testosterone, all of which are involved directly or indirectly in promoting cardiovascular disease, say the authors.

But they conclude: “[Our] findings suggest that vertex baldness is more closely associated with systemic atherosclerosis than with frontal baldness. Thus, cardiovascular risk factors should be reviewed carefully in men with vertex baldness, especially younger men” who should “probably be encouraged to improve their cardiovascular risk profile.”

The most-read article in February was Al-Shafaee and colleagues’ study of the mistreatment of clinical interns in Oman. Doyle et al‘s  systematic review of the links between patient experience and clinical safety – originally published in January – remains popular,  and Ubbink et al’s newly published scoping review on evidence-based practice was the third most-read.

Ward et al‘s paper on new drug launches in the UK received plenty of media attention, including articles by the BBC.

Most read figures are based on pdf downloads and full text views. Abstract views are excluded.

The widely held belief that the UK supply of innovative new medicines has conspicuously dwindled in recent decades, is not borne out by the evidence, reveals research published in BMJ Open.

In fact, the reverse may be true, the figures suggest.

The prevailing view is that pharmaceutical industry innovation has been in decline, with fewer new drugs launched in recent decades than before—despite more cash being pumped into research and development—prompting a good deal of hand-wringing, say the authors.

They wanted to find out whether this view was justified, by looking at all new medicines added every year to the prescribing and dispensing drugs bible, the British National Formulary, or BNF for short, over 30 years. The BNF is updated every six months.

After the US, the UK is the next largest source of new drug development, generating more than 10% of all new medicines around the globe.

All new synthetic chemical entities and new biological drugs, such as vaccines, blood products, and gene therapies were included, based on their first appearance in the BNF between 1982 and 2011.

New products covered modifications of existing drugs as well as radical breakthrough treatments. Different doses and formulations containing the same active ingredients were only counted once, and generic versions of brand drugs were excluded.

There was no significant linear trend pointing to a decline in the number of new drugs introduced into the UK over that period, which averaged just under 24 a year.

But the authors did find a pattern of peaks and troughs, with dips invariably followed by a surge in new arrivals.

After a dip in the mid-1980s, with around a dozen new drugs coming on to the market between 1985 and 1987, new arrivals increased every year, peaking at 34 in 1997.

This peak was again followed by a dip, with around 20 new drugs a year between 2003 and 2006, followed by a further peak in 2010.

And extending the timeline back to the 1970s indicated an overall slight but significant increase of 0.16 new drugs every year between 1971 and 2011, “contradicting the widely held view that the number of new medicines being launched is declining,” say the authors.

They conclude that the perceived “innovation dip” is a product of the time periods studied in the past.

“Although there was indeed a dip in new drug introductions during the decade from 1997 to 2006, this was largely an artefact of a peak in 1997, which was itself preceded by an unusually low number of launches in 1985-87,” they write. “Additionally, the peak number of new drugs added to the BNF in 1997 was matched in 2010.”

The authors point out that their study does not distinguish between varying degrees of innovation, and launches are not the only indicator of pharmaceutical industry health. But theirs is the most up to date UK study of new launch trends, they say.

Nevertheless, they do sound a note of caution: the costs of drug development have soared. and the time taken to bring a new drug to market has risen from 3 years in 1960 to 12 in 2000.

The most read article in January was Doyle et al‘s recently published systematic review of the links between patient experience and clinical safety. Bellis et al‘s much discussed paper on rock star mortality was in second place, followed by Kripke et al‘s study of hypnotics and mortality, originally published almost a year ago.

Newly published papers in the top ten include Mahtani and colleagues’ analysis of whether the London Olympics will increase participation in sport and Rowlingson et al‘s mapping study of English GP prescribing data.

Most read figures are based on pdf downloads and full text views. Abstract views are excluded.