“Study shows girls with absent fathers more likely to develop depression," the Mail Online has revealed.

It reports on a large UK study that found that girls whose biological fathers were absent during the first five years of their childhood had an increased risk of symptoms of depression. No increase in risk was found for girls whose fathers were absent later in childhood, and no increase in risk was found for boys with absent fathers.

Researchers collected information regarding the physical absence of the biological father during childhood, as well as information on depression symptoms when the child was 14. They assessed whether there was any association between these factors.

During their analysis, the researchers took into account several factors that may influence the link, such as family characteristics. However, despite the researchers' efforts to take these variables into account, the reasons why a father may be absent from the family home can be incredibly complicated. This means we can’t be sure whether other factors have produced the association between absent fathers and depression in girls.

Where did the story come from?

The study was carried out by researchers from the University of Bristol and was funded by the UK Medical Research Council, the Wellcome Trust and the University of Bristol.

The study was published in the peer-reviewed journal Psychological Medicine.

Media coverage of this research was broadly accurate, though neither ITV nor Mail Online outlined any of the study’s limitations.

What kind of research was this?

This was an analysis of data from a prospective cohort study called the Avon Longitudinal Study of Parents and Children. This is a study that has been ongoing since the 1990s that assesses influences on the health and development of children.

The researchers were interested in the potential link between the absence of the biological father in early childhood and the risk of mental health problems. They were specifically interested in symptoms of depression that were not necessarily severe enough to be considered clinical depression.

As a prospective cohort study, this research is less likely to be affected by certain types of bias, especially recall bias. It was important that the researchers collected data on the effect of family factors on the children’s mental health at the time, rather than at a later date, to help ensure the information was accurate. Prospective studies allow for this.

What did the research involve?

The researchers measured two main factors:

• absence of the biological father during childhood
• experience of depressive symptoms during the teenage years

To measure parental absence, the researchers used questionnaires, filled out by the children’s mothers regularly throughout the children’s lives. These questionnaires asked whether the ‘present live-in father-figure is the natural father of the child and, if not, how old the child was when the natural father stopped living with the family’. This information was used to divide the children into three groups:

• biological father present
• biological father not present during the first five years of life (during early childhood)
• biological father not present from age 5 to 10 (during middle childhood)

To assess the teenagers’ experiences of depressive symptoms, the researchers asked the study participants to complete a 13-item questionnaire when they were approximately 14 years old. This asked about the presence of certain symptoms over the previous two weeks. The questionnaire is reported to be a reliable and valid measure of depression in children. Children scoring 11 or higher on this questionnaire were considered to have high levels of depressive symptoms. This is not the same as being diagnosed with depression, however.

The researchers then analysed the data, comparing the risk of having high levels of depressive symptoms among children whose biological father left during early or middle childhood to the risk in children whose fathers were still living with them. These analyses were adjusted for several factors (confounders) that could be linked to both the absence of the father and depressive symptoms, including:

• socioeconomic status (including home or car ownership, major financial problems, family size and parents’ jobs)
• mother’s characteristics (including having a child before the age of 20, experiencing depression during pregnancy), and
• any parental conflict between the mother and her current partner

Separate analyses were carried out for boys and girls, to determine whether the child’s gender had any impact on the relationship between father’s absence and depressive risk.

What were the basic results?

There were approximately 14,500 children in the original cohort study, approximately 11,000 of whom had data available on the presence or absence of their biological father. Among these children, approximately 6,000 had available data regarding depressive symptoms at age 14.

Overall, girls reported higher levels of depressive symptoms than boys, regardless of whether their father lived with them or not – a trend that has also been found in previous studies.

Girls

The study included:

• 374 girls whose father left during early childhood, 87 (23.3%) of whom had high depressive symptoms at age 14
• 193 girls whose father left during middle childhood, 27 (14.0%) of whom had high depressive symptoms at age 14
• 2,295 girls whose father was present throughout childhood, 332 (14.5%) of whom had high depressive symptoms at age 14

Boys

The study included:

• 357 boys whose father left during early childhood, 30 (8.4%) of whom had high depressive symptoms at age 14
• 185 boys whose father left during middle childhood, 17 (9.2%) of whom had high depressive symptoms at age 14
• 2,227 boys whose father was present throughout childhood, 166 (7.4%) of whom had high depressive symptoms at age 14

When assessing the association between the absence of the father in early childhood and teenage depressive symptoms, researchers found that:

• Girls with absent fathers during early childhood had a 53% greater chance of experiencing high levels of depressive symptoms compared with girls with fathers present during this time (odds ratio  [OR] 1.53, 95% confidence interval [CI] 1.07 to 2.21).
• Boys with absent fathers were no more likely to report high levels of depressive symptoms at age 14 than boys whose fathers were present during early childhood (OR 1.08, 95% CI 0.65 to 1.79).

There was no significant association between middle childhood father absence and teenage depressive symptoms.

How did the researchers interpret the results?

The researchers concluded that “father absence in early childhood increases risk for adolescent depressive symptoms, particularly in girls”.

Conclusion

This large prospective cohort study suggests that there is a link between a father’s absence during the first few years of life and a girl’s risk of experiencing depressive symptoms.

This study has several strengths, including its large sample size, its long-term follow-up and prospective collection of data for the analyses. It also attempted to consider confounding variables during the analysis and was based in the UK, which helps to ensure that the results are applicable here.

There are some limitations, however, that should be taken into account, including the following.

• Only a third of the original cohort was analysed due to missing data on key factors. It is unclear to what extent those included differed from the entire population-based cohort. The researchers report that drop-outs were more likely among participants in lower socioeconomic groups. This factor is linked to both parental absence and depressive symptoms, so it could reduce the validity of the results and how much we can infer from them.
• The adjusted analyses further reduced the available sample size due to missing data on confounding factors, and the researchers suggest that this may have resulted in loss of statistical power to detect an effect.
• Several potential confounders were not included in the analysis, and could have influenced the results. The study authors report some of these potential confounders (quality of parent-child relationship, the father’s involvement in the child’s life regardless of whether he lived in the same house).
• The questionnaire used to assess depressive symptoms is not a measure of clinical depression. A high score on this questionnaire does not indicate that the child has or will develop a diagnosable depressive disorder.

Overall, this study suggests that early childhood family environments may play an important role in the mental health of children. At this stage we don’t know what accounts for the study’s results, and the researchers say that this should inspire future research into the possible biological and psychological mechanisms underpinning this relationship.

Depression is one of the most common mental health conditions, yet there is very little good quality evidence about how to prevent people developing depression. Research that gives us insight into the factors that increase children’s likelihood of developing depression would be invaluable.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

A quarter of young girls with absent fathers 'grow into depressed teenagers': Researchers say boys cope better with parental separation. Mail Online, May 15 2013

Study shows girls with absent fathers more likely to develop depression. ITV News, May 15 2013

Links To Science

Culpin I, Heron J, Araya R, et al. Father absence and depressive symptoms in adolescence: findings from a UK cohort. Psychological Medicine. Published online May 14 2013

"Depressed people are out of sync with the rest of the world because their body clocks are broken," reports the Mail Online website, while The Independent claims that depressed people live in a "different time zone".

The story comes from a study that looked at the activity of genes thought to be involved in regulating the body's internal clock – the innate sense that most people have of the changes over a 24-hour day to night cycle (circadian rhythms).

Researchers did a detailed study of gene expression, the effect that certain proteins contained inside individual genes have on genetic activities inside the body.

The study involved examining brain tissue taken from people who donated their brains to science after their deaths. Of the sample, 55 people had no history of psychiatric illness, while 34 patients had a history of severe depression (major depressive disorder, or MDD).

Researchers found that the gene activity associated with regulating circadian rhythms was much weaker, and often disrupted, in the brains of patients who had MDD.

These results possibly present, as philosophers put it, a "causality dilemma" (a chicken and egg problem) – does depression lead to a disrupted body clock, or does a disrupted body clock make people vulnerable to depression?

It is too early to say what help these findings may be in the understanding and treatment of MDD.

Where did the story come from?

The study was carried out by researchers from the University of Michigan, the University of California, Weill Cornell Medical College, Stanford University and the HudsonAlpha Institute for Biotechnology, and was supported by the Pritzker Neuropsychiatric Disorders Research Fund.

It was published in the peer-reviewed Proceedings of the National Academy of Sciences.

Both the Mail Online and The Independent covered the research uncritically. Given the specialised nature of this research, it's not surprising that both of the news stories appeared to be strongly based on an accompanying press release and were not a critical appraisal of the study itself.

What kind of research was this?

This was laboratory research using donated post-mortem brains. In it, researchers analysed in detail the gene expression of certain genes thought to be associated with circadian rhythm regulation at the time of death.

The authors point out that a common symptom of major depressive disorder is the disruption of circadian patterns, which can trigger symptoms of insomnia as well as excessive daytime sleepiness and fatigue (feeling tired all the time). However, to date there is no direct evidence of "circadian clock dysregulation" in the brains of patients with major depressive disorder.

What did the research involve?

Researchers used human brain tissue taken from a US donor programme with the consent of next of kin. They also took information from medical records, medical examiners and interviews with relatives to record the donors' previous physical health, medication use, psychiatric problems, substance use and details of death.

This was done in order to assess whether donors had a major depressive disorder, a severe form of depression that has a significant impact on day-to-day living.

They also assessed whether physiological stress at the time of death would have had an effect on gene expression, and took account of this potential confounding factor.

Researchers analysed the brain tissue of 55 donors with no history of psychiatric or neurological illness and 34 patients with major depressive disorder. Using specialist techniques called DNA microarray, they measured the expression of genes thought to be associated with regulating circadian rhythms in different areas of the brain.

They used the control group to build a detailed picture of circadian gene expression in brain tissue and compared the results with those found in the brains of people with MDD. They also used the rise and fall of the top 100 "cyclic" genes in 60 of the donors to predict the time of death in all the others, both cases and controls. 

What were the basic results?

In the brain tissue from donors without major depressive disorder, they found that the activity of "circadian" genes at certain times of the day and night was consistent with data derived from other diurnal (day-active) mammals. More than 100 genes showed "consistent cyclic patterns" over six brain regions.

However, in the brains of patients with MDD gene expression of cyclic patterns was far weaker and more disrupted, with the patients' day pattern of gene activity often resembling a night pattern.

They found that predictions of time of death were more accurate among controls than for those with MDD.

How did the researchers interpret the results?

The researchers say the results provide convincing evidence that there is a "rhythmic rise and fall" in the activity of hundreds of genes in the human brain associated with regulating the day/night cycle. There is also evidence that the activity of genes associated with circadian rhythms is abnormal in people with MDD.

The study identifies hundreds of genes in the human brain that are likely to be involved in the sleep/wake cycle. The researchers conclude that daily rhythms in these genes are "profoundly dysregulated" in MDD. They say the results pave the way for the identification of new biomarkers and treatments for mood disorders.

Conclusion

This study is of interest, but at the moment it has little bearing on our understanding and treatment of depression. It could lead to new insights and treatments in the future, but there is no guarantee that this will be the case.

Also, as the authors point out, gene activity can result from many factors, including disease and drug history. In particular, it should be pointed out that:

• the researchers relied on only 55 patients to build a "normal" picture of genetic expression associated with the sleep/wake cycle
• it is not clear whether those in the MDD group had all been formally diagnosed with MDD or how long they had had depression, and it is possible there were errors in the classification of patients either with or without MDD

In conclusion, it is too early to say whether this study's findings might help in the understanding and treatment of major depressive disorders.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

New forensic technique for estimating time of death by checking internal clock of the human brain. The Independent, May 13 2013

Out of sync with the world: Depressed people suffer with 'broken body clocks'. Mail Online, May 14 2013

Links To Science

Zi JZ, Bunney BG, Meng F, et al. Circadian patterns of gene expression in the human brain and disruption in major depressive disorder. PNAS. Published online May 13 2013

‘Research has found emotional eaters tend to eat more when happy’, reports the Mail Online website.

The news is based on a small study looking at whether experimentally altering mood has an effect on the amount of calories a person eats.

The researchers examined the effects on what they describe as ‘emotional eaters’ – people who reported using food as a coping mechanism for emotions.

A group of 86 students, who said they were either emotional or non-emotional eaters, were shown TV and movie clips to evoke either a positive, negative or neutral mood. The researchers then assessed how much the students ate when provided with bowls of crisps and chocolate, as well as assessing their change in mood.

Emotional eaters who were shown the positive mood-inducing scenes significantly increased their food intake compared to emotional eaters shown the neutral mood-inducing scenes. However, the negative mood-inducing scenes had no effect on food intake of emotional or non-emotional students.

The common assumption is that emotional eaters eat more when in a negative mood, but this study provides very limited evidence to suggest that this may not always be the case.

However, because this experiment was based in a laboratory and researchers did not measure how hungry people were, even this finding should be viewed with caution. As ever, more and better research is needed if people with eating disorders or weight problems are to be helped effectively.

Where did the story come from?

The study was carried out by researchers from Maastricht University in The Netherlands and was funded by the Netherlands Organisation for Scientific Research. It was published in the peer-reviewed journal, Appetite.

The story was picked up by the Mail Online website and it was covered appropriately, although the limitations of the study could have been described in more detail.

What kind of research was this?

This was a laboratory study looking at the effect of experimentally influencing mood changes in a group of students reported to be emotional or non-emotional eaters, and then looking at the effect on their food and calorie intake.

The researchers say emotional eaters are thought to increase their food intake in response to negative emotions, but little is known about the effect of positive emotions on their food intake. Meanwhile, non-emotional eaters are not believed to change their intake levels in response to emotions, and they might even restrict food intake in response.

The main limitation of this research is that a study of a small, select population sample under experimental conditions can only provide very limited indications about the possible influence emotions may have upon the eating patterns of different people in daily life.

For example, if you thought that researchers could be measuring how much you were eating it could make you, perhaps unconsciously, reluctant to eat as much as you normally would. Alternatively, being in this type of study could make you nervous, leading you to eat more than you normally would.

What did the research involve?

The researchers recruited 86 psychology students in their second year at Maastricht University in the Netherlands who received credit points for their participation. The students were predominantly female (75%) and had an average age of 21.6 years (range 19 to 43).

The students answered a series of questionnaires to assess their mental health and eating behaviours. Emotional eating was assessed using a questionnaire called the Dutch Eating Behaviour Questionnaire (DEBQ). Students were asked, ‘Do you have a desire to eat when you’re feeling lonely?’ and provided answers on a five-point Likert scale that ranged from ‘never’ to ‘very often’.

The researchers then carried out a series of experiments in a laboratory setting that aimed to change the student’s mood. Students were randomly allocated to view clips from television or films that aimed to evoke either a positive, negative or neutral mood:

• 28 students were shown two clips to evoke a positive mood. Firstly, they were shown a scene from the television series Mr Bean (which showed Mr Bean struggling to copy answers from his neighbour during an exam). The second clip was taken from the movie ‘When Harry Met Sally’ which showed the famous scene where Meg Ryan’s character simulates an orgasm in front of other diners in a restaurant.
• 28 students were shown one negative clip from the film ‘The Green Mile’, which showed an innocent man being executed.
• 30 students were shown part of a documentary about fishing to evoke a neutral mood.

The students were told to give in to the emotions the clips evoked, and were presented with bowls containing 191g of chocolate (white, milk and dark, equivalent to 1,000 kcal), 225g of salted crisps (1,229 kcal) and 225g of ketchup crisps (1,217 kcal). The bowls were weighed before and after the experiment to determine the amount of food eaten and calorie intake. 

The students were asked to assess their mood using a visual analogue scale (this is essentially a straight line – where the far left of the line represents poor mood and the far right represents very good mood) at five points during the experiment:

• before the experiment began
• immediately after watching the television or movie scenes
• 5 minutes after the experiment
• 10 minutes after the experiment
• 15 minutes after the experiment

The students were told when entering the laboratory that they were taking part in an experiment on the effect of movie clips on taste perception.

The researchers analysed their results using validated methods and adjusted the results for gender, body mass index (BMI), external eating and dietary restraint as assessed by the DEBQ, and negative mood as assessed by the Positive and Negative Affect Schedule (PANAS).

What were the basic results?

Overall, there was no significant difference between emotional eaters eating more than non-emotional eaters who were shown positive, negative or neutral clips.

When looking specifically at only the emotional eaters:

• those shown the positive mood-inducing scenes significantly increased their intake of food compared to those shown the neutral mood-inducing scenes
• there was no difference in food intake between students shown negative mood-inducing scenes and those shown neutral or positive mood-inducing scenes

How did the researchers interpret the results?

The researchers concluded that self-reported emotional eaters respond in a different way to emotions than non-emotional eaters. They say that emotional eaters ate more in a positive mood compared to a neutral mood, whereas non-emotional eaters ate about the same amount in both conditions.

In discussing the results, the researchers say the findings could be of value for the treatment of obesity.

Conclusion

Overall, this small study provides very limited evidence to suggest emotional eaters eat more when feeling in a positive mood. There are several limitations to this study, some of which are noted by the researchers. These include the facts that:

• the laboratory setting may not be an appropriate setting to test emotional eating with different mood feelings. It is possible that students felt uncomfortable in this setting and limited their food intake as they were being watched
• the students were told they were partaking in an experiment of taste perceptions, so may have been inclined to eat more than they normally would have because of what they were told the study was looking at
• no hunger measurements were taken during the study and how hungry each student was could have greatly affected the results
• there was no group included in the study that did not eat, so it is not possible to say from the findings that the changes in mood were due to food intake
• all of the participants were students, so findings may not be the same as if the same experiments were carried out in different groups who report being emotional eaters

To draw firmer conclusions about the effects of mood on emotional eating, larger studies of different groups are required that carry out experiments in more natural environments. 

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

Forget comfort eating - could happiness be the reason you're piling on the pounds? Mail Online, May 13 2013

Links To Science

Bongers P, Jansen A, Havvermans R, et al. Happy eating: The underestimated role of overeating in a positive mood. Appetite. Published online April 10 2013

'Redheads are at increased risk of skin cancer even if they don't spend time in the sun,' is the headline on the Mail Online website.

The story refers to a discussion piece in a journal that outlines theories about the results of some animal experiments. This research involved mice genetically engineered to have red fur and predisposed to develop melanoma.

Although exposure to ultraviolet (UV) light is known to be a major risk factor for melanomas, the researchers found that genetically engineered mice with red fur still had a high risk of developing melanomas even without UV exposure.

The article discusses potential explanations for why this could be the case, and these theories now need to be tested to see if they are correct.

It is not yet clear how well these animal studies represent what happens in people with red hair. It would be very difficult to test this directly, as keeping people completely away from sunlight would be impractical and potentially unethical.

UV light exposure is known to increase the risk of melanoma in redheads and non-redheads alike. It is important that people with red hair should continue to use sensible precautions to avoid excessive UV exposure and sunburn, despite this news.

Where did the story come from?

The article was written by researchers from the Cutaneous Biology Research Center at Massachusetts General Hospital in the US.

No sources of funding for the article were reported. It was published as an "Ideas and Speculations" article in the journal BioEssays. These pieces are described as "creative thinking and predictions on open questions and recent developments in biology".

The article has been peer-reviewed.

The news is based on an article by researchers that presents possible explanations for their previous finding that genetically engineered mice with red fur and a predisposition to melanoma develop this cancer even without UV exposure.

Some of the Mail Online reporting suggests that the findings of this research are more conclusive than is possible to say at this stage: "Scientists have discovered that the production of red hair pigment causes an increased risk of melanoma".

However, the BioEssays article was only presenting possible explanations for observations from animal experiments. It was not claiming to have definitive proof that these findings apply to humans.

What kind of article was this?

This was an article discussing the potential link between the red pigment in red hair and skin cancer.

People with red hair and fair skin are known to be at greater risk of getting melanoma, the least common but most serious form of skin cancer, which is responsible for around two thousand deaths a year in the UK.

In general, it is thought that redheads' pale skin makes them more susceptible to UV damage from the sun's rays.

However, the authors of the article say that a recent study from their lab suggests that the pigment that causes hair to turn red (pheomelanin) could itself be linked to the increased risk of cancer, even without UV exposure.

In their article, the authors discuss two possible ways in which the red pigment in red hair might increase the risk of cancer. These preliminary ideas – or hypotheses – are based on previous research and a general understanding of human and cancer biology.

A hypothesis is a possible explanation of why something that researchers have observed might happen. Researchers design experiments to test whether their hypothesis is correct. This process is fundamental to the scientific method.

What did the article say?

The researchers first describe how the red colour in red hair is made, and discuss the results of their recent study before going on to present their hypotheses.

Specific cells in the skin called melanocytes make two kinds of pigment – a brown pigment called eumelanin and a red-orange pigment called pheomelanin. A biochemical process within cells determines how much of each pigment is made.

This process involves a protein called MC1R, which influences the switch between the production of these pigments based on the strength of the signal it sends to the cell and whether the cell has enough of the amino acid cysteine.

In redheads, variations in the gene for the MC1R protein means that it sends weak signals. This means that the cells' stores of cysteine are usually enough for it to favour producing the red/orange pigment pheomelanin.

The researchers recently carried out a study where they introduced a genetic mutation commonly found in melanoma cells into the melanocytes of mice. When they also introduced a genetic mutation into these mice that inactivated the MC1R protein, the mice had red fur and developed melanoma, even without UV exposure. If they introduced another genetic mutation that stopped pigment being made altogether, the mice were albino but they did not develop melanoma.

This led the researchers to suspect that the red pigment pheomelanin could be itself increasing the risk of melanoma. Their research also found that the mice with red fur had more damage to their skin cell DNA caused by very reactive chemicals called free radicals. Free radicals can cause damage to cells at a molecular level.

The researchers do not yet know how the red pigment might be linked with the free radical DNA damage that can increase the risk of melanoma. However, they have presented two hypotheses:

The first hypothesis

The researchers' first hypothesis was that the red pigment itself might generate more free radicals, and that these cause DNA damage that could lead to melanoma. They say that the red pigment is already known to make free radicals when it is exposed to UVA light, but it may be able to do this without UVA light. These free radicals could potentially:

• damage DNA directly
• damage its building blocks, or
• use up the cell's stores of antioxidants, making it more vulnerable to damage by other free radicals

The researchers also discuss in detail the biochemical ways in which the red pigment might generate free radicals.

The second hypothesis

The second hypothesis was that the process of making the red pigment might use up the cell's stores of antioxidants, rather than the red pigment itself. This might make the cells more vulnerable to damage by other free radicals.

They say that the amino acid cysteine used in making the red pigment is also found in the most important antioxidant in the cell, glutathione. If cysteine is used to make the red pigment, this might reduce the cell's ability to make this antioxidant.

The researchers report that red-haired wild boars have been found to have less glutathione in their muscles. However, they acknowledge that it is not possible to say from this whether there is less glutathione due to free radicals from the red pigment itself or the making of the red pigment.

What were the researchers' conclusions?

The researchers presented two hypotheses that could explain how the red skin and hair pigment pheomelanin could increase the risk of the skin cancer melanoma.

They say that their two proposed methods could both be occurring, and that more research could help identify how redheads can reduce their risk of melanoma.

Conclusion

The researchers' article discusses potential ways in which the red pigment found in the cells of people with red hair might increase the risk of melanoma, the most serious form of skin cancer. It is not a standard report of a research study, but the authors put forward potential explanations for their previous research findings. These now need to be tested to see if they are correct.

The researchers' previous research found that mice genetically engineered to be predisposed to melanoma and red fur developed melanomas even without UV exposure. It is not clear to what extent these genetically engineered mice represent what happens in humans.

It would be very challenging to test this – keeping people completely away from UV light would not be feasible or ethical, as we need some sun exposure to make vitamin D, which is needed to make and maintain strong bones. For this reason, research in mice can be very helpful.

It is important that redheads do not take this news as a reason not to protect themselves from the effects of the sun. We already know that UV light exposure increases the risk of melanoma in people regardless of hair colour. People with red hair should continue to use sensible precautions to avoid excessive UV exposure and sunburn.

Read more about reducing your melanoma risk. 

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

Redheads are at increased risk of skin cancer even if they don't spend time in the sun. Mail Online, May 10 2013

Links To Science

Morgan AM, Lo J, Fisher DE. How does pheomelanin synthesis contribute to melanomagenesis? BioEssays. Published online May 7 2013

‘Children's exposure to traffic pollution could…lead to diabetes' BBC News explains, reporting on a German study.

The study included around 400 children aged 10. Researchers looked at measures of air pollution and proximity to the nearest road at the address each child had lived as a baby.

They also measured each child’s blood sugar and insulin levels.

The second measurement allowed them to calculate each child’s level of insulin resistance – to what extent the cells of the body fail to respond to the hormone insulin (which the body uses to convert blood sugar into energy).

Once insulin resistance reaches a certain level, the symptoms of type 2 diabetes can develop.

The researchers found an association between exposure to air pollution and increased levels of insulin resistance.

However, an association is not the same as proof of a direct causal effect. Living near a busy road would usually imply that a child lives in an urban environment. So there could be a range of environmental factors, other than air pollution, affecting levels of insulin resistance (as well as a wide-range of other possible individual genetic and health-related factors).

The study also does not tell us whether any insulin resistance measured in the child actually had any clinical significance and would lead to a child developing diabetes in later life.

Due to these limitations, further studies in other population samples would be useful.

Where did the story come from?

The study was carried out by researchers from German Center for Diabetes Research and other institutions in Germany, and was funded by German Federal Ministry of Education and Research, and The European Community’s Seventh Framework Programme.

The study was published in the peer-reviewed medical journal Diabetologia.

The quality of the reporting on the study in the UK media is mixed. The BBC News headline gives an accurate representation of the current study as it includes the all-important word ‘may’. However, the Mail Online’s headline linking air pollution to a child’s risk of developing diabetes may be misleading.

This study has many limitations, not least, that increased levels of insulin resistance in childhood, while a risk factor, is not a guarantee that a child will grow up to develop type 2 diabetes.

Also, the association between childhood and diabetes may confuse some readers into thinking that the study was looking at type 1 diabetes – the form of the condition that normally begins in childhood and where the body’s own immune system destroys the insulin-producing cells, so the person is not able to produce any insulin at all.

What kind of research was this?

This was a cohort study looking at whether there was an association between air pollution and insulin resistance.

The researchers say that previous research has shown that traffic and air pollution may increase the risk of diseases affecting the lungs and cardiovascular system.

This is speculated to be due to exposure to pollution that may trigger oxidative stress (a disruption in the body’s ability to repair cellular damage). Pollution could also lead to low levels of inflammation in certain cells of the immune system and those lining the blood vessels.

Animal studies have also suggested that pollution may make cells of the body more resistant to the action of insulin – the hormone released from the pancreas that helps the body to make use of the glucose in the blood.

The researchers say that no study has yet looked at whether traffic-related air pollution can lead to insulin resistance in school-aged children. This German cohort study aimed to look at the relationship between particulate matter in the air and proximity to the nearest road at the child’s birth address, and the child’s insulin resistance when they reached the age of 10.

The limitations of such a study include it being difficult to conclude that the air pollution at the birth address has directly caused the child’s insulin resistance at age 10.

There may be many other genetic, environmental and health-related factors involved.

The study also does not tell us whether any insulin resistance measured in the child has any clinical significance, and whether it is related to later development of type 2 diabetes in adult life.

What did the research involve?

The researchers included sub-groups of 10-year-old children taking part in two separate birth cohorts in Munich, South Germany and Wesel, West Germany:

• The German Infant Study enrolled almost 6,000 healthy newborns and was a trial looking at the effect of a hypoallergenic infant formulae on a child’s risk of allergy (in addition to looking at other environmental and genetic influences).
• The Lifestyle-Related Factors study included just over 3,000 healthy newborns and was an observational study looking at the effect of lifestyle factors on the child’s immune system and risk of allergies.

The present study included 397 children randomly sampled from these two cohorts (though 82% came from the Munich cohort) who had blood samples taken for insulin and glucose measurement at age 10, and who had information available for air pollution exposure at the time they were born.

To measure pollution exposure at the birth address, the researchers used models to estimate levels of:

• nitrogen dioxide (N02)
• particulate matter of less than 2.5 micrometres in diameter
• particulate matter of less than 10 micrometres in diameter

Particulate matter is the term for a mixture of solid particles and liquid droplets found in the air.

Measurements were taken at selected monitoring sites on three occasions over 14 consecutive days, and in different seasons.

When conducting their analyses, factors taken into account at each monitoring site were location, surrounding land use, population density and traffic patterns.

Other factors taken into account that could have an influence on the results (confounders) related to the individual child included:

• parental education (used as an indicator of socioeconomic status)
• exposure to second-hand smoke
• height and weight at age 10
• whether they had started to go through puberty

What were the basic results?

There were no differences between children in the two cohorts, except that those from Wesel were more likely to have been exposed to second-hand smoke and to be of lower socioeconomic status. Pollutant levels were also higher in Wesel than Munich.

After adjustment for all potential study-centre and child-related confounding factors, each two-point standard deviation increase in nitrogen dioxide levels was associated with a 15.8% increase in insulin resistance (95% confidence interval (CI) 3.8 to 29.1).

Each two-point standard deviation increase in particulate matter of less than 10 micrometres in diameter, was associated with a 17.5% increase in insulin resistance (95% CI 1.9 to 35.6). There was no significant association with particulate matter of less than 2.5 micrometres in diameter.

Distance to the nearest road, as would be expected, was significantly associated with pollutant levels (shorter distance equalled higher levels of nitrogen dioxide and particulate matter). Shorter distance to the road was also associated with increased insulin resistance (each 500 metre decrease in distance to road increased insulin resistance by 6.7%, 95% CI 0.3 to 13.5).

The researchers found that the link between pollution levels and insulin resistance was stronger in children who had not moved from their birth address by age 10.

How did the researchers interpret the results?

The researchers conclude that traffic-related air pollution may increase the risk of insulin resistance in children. They say that the associations observed may have important public health implications despite the small effect seen.

Conclusion

This German study looked at the relationship between air pollution and proximity to the nearest road at the child’s birth address, and the child’s insulin resistance when they were aged 10. Though links were found between increasing levels of nitrogen dioxide and levels of particles less than 10 micrometres in diameter and increasing insulin levels at age 10, there are important limitations to bear in mind:

• Though the researchers have attempted to adjust for many potential confounders, it is difficult to conclude that the air pollution at the birth address has directly caused the child’s insulin resistance at age 10, when there may be many other genetic, environmental and health-related factors involved.
• The confidence intervals around the increase in insulin resistance with each incremental increase in pollutant levels are very wide. For example, each increase in particles of less than 10 micrometres was associated with a 17.5% increase in insulin resistance, but the actual increase could lie anywhere between 1.9% and 35.6%. This means we can have less confidence in the reliability of these estimates.
• The study does not tell us whether any insulin resistance measured in the child has any clinical significance, and whether it will be related to higher risk of developing type 2 diabetes in adult life.
• Also, as mentioned above, the news headlines should not be wrongly interpreted to mean that a child has increased risk of developing type 1 diabetes – the type that readers may associate with start in childhood.  
• Lastly, the results are based on only a relatively small sample of children from two regions in Germany. Studies of much larger samples from different countries would give more weight to any observations.

Overall, this study cannot prove that air pollution increases a child’s risk of developing diabetes, only that there may be an association with insulin resistance.

As it is unlikely that we are going to live in a world free from air pollution anytime soon, the most effective way of reducing your child’s diabetes risk is to encourage them to take plenty of exercise and eat a healthy diet. These types of good habits in childhood often carry on into adulthood meaning that your child is more likely to maintain a healthy weight – a proven method of reducing type 2 diabetes risk.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

Diabetes: dirty air 'may raise' insulin resistance risk. BBC News, May 10 2013

Why living near a busy road could be dangerous for your child's health: Traffic pollution linked to diabetes risk in children. Mail Online, May 10 2013

Air pollution could raise diabetes risk. The Daily Telegraph, May 10 2013

Links To Science

Thiering E, Cyrys J, Kratzsch J, et al.  Long-term exposure to traffic-related air pollution and insulin resistance in children: results from the GINIplus and LISAplus birth cohorts (PDF, 278KB). Published online May 9 2013

"Middle class professionals… are the country’s biggest problem drinkers," is the stark and somewhat misleading claim in The Daily Telegraph, with similar claims appearing across the UK media.

The story is based on a study looking at just 49 ‘white collar’ people’s attitudes towards alcohol consumption. The study involved interviewing five small groups in a ‘focus group’ setting.

Researchers found that among these small groups:

• problem drinking was something that was seen to happen to other people – such as teenagers in city centres or binge drinkers in pubs
• if regular alcohol consumption did not significantly disrupt day-to-day functioning (such as in work or parenting skills) or lower social standards, then it was acceptable and harm-free
• regular ‘controlled’ drinking at home (for example as a way to relax), was also acceptable and harm-free 

It is important to note the study was very small and these findings may not be applicable to other countries or cultures. However, the attitudes reported suggest that key messages of some public health campaigns about reducing harm from alcohol are going unheard or being ignored.

It is not just binge drinking that can damage your body; regularly drinking above the recommended limits – whatever the social context – can also be harmful.

Where did the story come from?

The study was carried out by researchers from Newcastle University and the University of Sunderland, UK and was funded by the Public Health NHS Directorate Stockton-on-Tees.

It was published in the peer-reviewed journal, BMC Public Health and made freely available to read on an open-access basis.

The story was picked up widely in the media. While the findings of the study were reported accurately, the tone of some of the reporting was slightly confusing.

It appears that some of the media fail to understand the nature and implication of this method of qualitative research. Such studies can provide useful insights into people’s attitudes and behaviours; however, they cannot provide hard statistical evidence. So headlines such as the Daily Express’s “Middle class ‘drink more than teens’” are misleading, as are sweeping statements like the Telegraph’s “Middle class professionals who drink at home are the country’s biggest problem drinkers”.

What kind of research was this?

This was a qualitative study looking at the drinking habits of a small number of adult ‘white collar workers’ in the UK. The study explored their views on alcohol use, how public health messages about alcohol are perceived, and the role alcohol plays in the workers' personal and professional lives.

The researchers say that little is known about white collar workers’ views on drinking alcohol.

Qualitative research uses individual in-depth interviews, focus groups or questionnaires to collect, analyse and interpret data on people’s behaviours and the reasons behind them. Typically, the number of participants is relatively small, but the transcripts from interviews and focus groups provide a large amount of data. Such studies report on meanings, concepts, definitions, metaphors, characteristics, symbols, and descriptions. As such, their conclusions can be more subjective than quantitative research, as questions are often exploratory and open-ended.

What did the research involve?

The researchers carried out interviews with 49 volunteers (17 male, 32 female) from five workplaces in the UK. The participants were aged between 21 and 55 years and were all working full time (at least 35 hours per week). To be included, participants had to be working in managerial, supervisory, clerical or other professional roles, referred to by the researchers as ‘white collar workers’.

Group interviews (focus groups) were carried out by the researchers at each of the five workplaces during lunch breaks. The five focus groups were made up of workers from:

• local government offices (focus groups one and two)
• a private sector chemical storage company (focus group three)
• a prison (focus group four)
• a tax office (focus group five)

The group interviews lasted between 45 and 75 minutes and were led by two researchers. The researchers used open-ended questions loosely based around four main themes related to drinking alcohol:

• lifestyle behaviours
• drinking at home
• variations in drinking during the week
• the effect of drinking on work

The researchers say areas of agreement and disagreement were explored with the participants, and that questions were continually adapted depending on the flow of conversation. The participants were informed that the purpose of the research was not to find out the quantity or frequency of alcohol consumption of the volunteers. The volunteers were given a £5 voucher and lunch for their time.

The researchers then used a particular technique called ‘constant comparison’ to analyse their results and grouped findings into themes relating to views of alcohol.

What were the basic results?

After analysing the focus group findings, the researchers reported three main themes.

Unacceptable or problem drinking

Unacceptable or problem drinking was perceived by the volunteers as being associated with long-term, heavy or binge drinking of ‘others’. The researchers reported that the participants highlighted ‘others’ as including young people, people with complex needs, and other stereotypes. The perception of excessive drinking was associated with looks and behaviour, rather than how much they have drunk. Personal drinking was viewed as a controlled choice rather than something they ‘need to do’.

Drinking at home

Drinking at home was considered normal, convenient and a socially acceptable form of relaxation from the responsibilities of work or parenting. Volunteers reported less drinking at ‘leisure premises’ such as a bar or pub, and driving was identified as the greatest factor influencing drinking behaviours. Drinking alcohol was considered part of everyday life and not something that interferes with other parts of life or causes harm.

Effect of drinking on functioning

Ability to function at work and act responsibly were key indicators of whether drinking was within acceptable limits. So, if a person was able to maintain employment in skilled jobs, they were therefore perceived to be drinking in a way that was not considered harmful. Despite awareness of guidelines for drinking, little notice was taken by participants and there was confusion about what a ‘unit’ was, the researchers report. Public health messages were also considered to have little or no personal relevance.

The researchers say that discussions indicated that the volunteers' reported alcohol use exceeded recommended guidelines for both the amount and how often drinking occurred. Interestingly, when the ill effects of alcohol were discussed, they were reported as only in relation to coping with a hangover and the loss of valuable time while feeling unwell. More subtle, insidious adverse effects such as gradual loss of liver function did not appear to occur to the volunteers.

Finally, lunchtime drinking at work was considered a 'thing of the past' and very much taboo.

How did the researchers interpret the results?

The researchers say that this study helps reveal the meanings attached to alcohol use by white collar workers and identifies resistance to public health messages. They say, "these findings suggest that current public health interventions have not been effective in engaging this group who are likely to drink at unhealthy levels but be highly resistant to reducing their alcohol consumption – especially as they do not consider their use to be problematic unless it impairs their capacity to fulfil responsibilities or function at work".

They conclude by saying, "future public health messages around alcohol should be less focused upon the crime and personal safety implications of irresponsible drinking and be more sensitive to the lifestyles and long-term health of the populations they target".

They add that further research is needed to identify which factors (other than driving) would engage white collar workers to change their views and drinking behaviours.

Conclusion

Overall, this research provides some early findings of so-called ‘white collar’ workers' views of drinking behaviours in the UK.

Although the study was very small, with only 49 volunteers' views analysed, it is useful in determining emerging themes, and the researchers do state that there was relative consistency across the five groups. The researchers also note that ‘strong personalities’ within the group may have influenced how the other participants responded.

Research among larger groups of white collar workers is needed to draw firmer conclusions about the drinking culture in the UK. It is worth noting that these findings may not be applicable to other countries or cultures. Ethnicity, cultural identity and religious beliefs of the participants were not reported, which may have influenced how participants responded to the questions.

One important final message to stress – and one that seems to have not been grasped by the volunteers in the study – is that it is not where you drink, why you drink, or who you drink with that matters. It is how much you drink.

Regularly exceeding the recommended daily amounts for alcohol can harm your health, regardless of whether it is with three litres of strong white cider on a park bench, or a bottle of Californian merlot in front of a DVD box set. 

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

Middle class wine drinkers 'think they know better than health experts'. The Daily Telegraph, May 8 2013

Middle classes are 'ignoring the dangers of using casual drinking to combat stress'. Daily Mail, May 9 2013

Alcohol abuse: Concern over middle-aged middle classes using booze to relieve stress. Daily Mirror, May 8 2013

Middle class ‘drink more than teens’. Daily Express, May 9 2013

Links To Science

Ling J, Smith KE, Wilson GB, et al. The ‘other’ in patterns of drinking: A qualitative study of attitudes towards alcohol use among professional, managerial and clerical workers. BMC Public Health. Published online October 23 2012

"Hungry shoppers 'buy more calories'," BBC News reports in a story based on a very small short-term study. The somewhat artificial study examined the effects of people skipping meals due to everything from busy lifestyles to intermittent diets such as the 5:2 diet.

These intentional or unintentional fasts may lead to unhealthy food choices being made at the shops. This research looked at whether being deprived of food for just a few hours has an effect on the types of food people opt for.

During a simulated shopping experience, the researchers found that people who were hungry selected more high-calorie foods than people who had just eaten a snack.

Similarly, people who went food shopping during times of the day when the researchers expected them to be hungry (late afternoon) purchased more high-calorie foods than people who shopped when the researchers thought they were less likely to be hungry (early afternoon).

However, no definitive conclusions can be drawn from these findings. The research had many limitations, including the fact that the first study was laboratory-based and laboratory findings may not reflect the real world. 

But it is common sense to grab a bite to eat before heading to the shops, and might be worth a try if you do find that shopping when hungry means you make less healthy food choices.

Where did the story come from?

The study was carried out by researchers from Cornell University in the US and was funded by the university. It was published in the peer-reviewed Journal of the American Medical Association (JAMA) Internal Medicine.

The BBC covered the study well, if slightly uncritically, as the study's limitations were not discussed.

What kind of research was this?

This research included two components (a laboratory study and a field study) designed to determine whether short-term food deprivation changes affect food shopping habits.

The researchers say that food deprivation has been shown to change how much food people buy, and fasting is known to alter how the brain reacts to certain foods. They were interested to know whether shopping while hungry also effects the types of food people purchase.

Laboratory and field studies can provide interesting information about how people may react in given situations, but they are prone to bias and confounding. These potential limitations should be kept in mind when considering the results of the study.

What did the research involve?

In the first part of the study, the researchers recruited 68 paid participants with ages ranging from 18 to 62 years. They were asked to avoid eating for five hours prior to the start of the experiment.

The participants were grouped together in sessions of six to 12 people. In half of these sessions, a plate of crackers was offered at the beginning of the experiment and participants were asked to eat enough of the crackers so that they were not hungry. The participants were not offered any food in the remaining sessions.

The groups then completed an experiment meant to simulate buying groceries online. The online store offered a mix of lower calorie foods (including fruits, vegetables and chicken breasts) and higher calorie foods (including sweets, salty snacks and red meat). The products were displayed without prices. The researchers recorded and compared the food choices of individuals who did not eat prior to the study with those who had been offered a snack.

The second study involved the observation of individuals in a more natural setting. The researchers tracked the food purchases of 82 people.

The first group were tracked during the early afternoon, or "low hunger hours" (between 13:00 and 16:00), when the researchers expected them to have had lunch and therefore not be hungry.

The second group was tracked during the early evening, or "high hunger hours" (16:00 to 19:00), when researchers thought they would have gone several hours without a meal.

The researchers characterised the food purchases as either high-calorie or low-calorie, and compared the number of foods that fell into each category between the two participant groups.

They statistically compared the number of low-calorie items, the number of high-calorie items, and the ratio of low- to high-calorie purchases between the groups.

What were the basic results?

The researchers found that participants in the hungry and not-hungry groups of the laboratory study chose a similar number of total items (approximately 14 in the hungry group versus 12 in the not-hungry group).

The two groups also chose similar numbers of low-calorie foods (approximately eight in both groups), but the hungry group selected significantly more high-calorie items (an average of nearly six, compared with four in the not-hungry group).

During the field study, the researchers found that participants in the evening group purchased fewer low-calorie items (approximately eight items) than the afternoon group (approximate average of 11 items). There was no statistical difference in the number of high-calorie foods purchased (approximately four in both groups).

The ratio of low- to high-calorie items (with a higher ratio indicating better food choices overall) was significantly higher in the early afternoon group (approximately four low-calorie items per each high calorie item) compared with the evening group (approximately 2.5 low-calorie items per each high-calorie choice).

How did the researchers interpret the results?

The researchers concluded that, "even short-term fasts can lead people to make more unhealthy food choices" by picking fewer low-calorie foods.

Conclusion

This study suggests that how hungry you are when you shop for food can have an impact on the food you choose.

This may not be too surprising for anyone who has made a quick trip to the shops while hungry and found themselves at the till with a basket full of crisps and biscuits, but no fruit or vegetables.

While interesting, the study has some limitations that should be noted:

• Both experiments were fairly small, with less than 100 people in each.
• A laboratory-based study such as the first experiment tends to be stronger the more it mimics the real world. A simulated online grocery shopping experience that removes item price as a contributing factor is less likely to mirror real-life decision making.
• The field study makes assumptions about hunger levels based on the time of day. This may not be a reliable manner in which to assess hunger – for instance, individuals shopping during the "low hunger hours" of 13:00 to 16:00 may have skipped lunch, while individuals in the "high hunger hours" may have had a late lunch, a snack, or an early dinner.
• Field studies are prone to confounding due to difficulties measuring and controlling for different factors that may also have an influence. It is not reported how the afternoon and evening shoppers differed, and it is possible that the relationship between the time of day and shopping choices was influenced by different participant characteristics, such as age, employment, education, or socioeconomic status, and not by hunger.

The researchers say that short-term fasting is fairly common and can arise from skipping a meal, either intentionally as part of a religious fast or in an effort to lose weight, or unintentionally due to chaotic work schedules.

However, given that it is pretty easy and low-risk to grab a snack before heading out to the supermarket, this may seem like a sensible thing to do – it may have the benefit of subtly changing the foods you buy and consume throughout the week.

If you are trying to lose weight or eat a healthier diet, it may be a good idea to plan your shopping in advance. Options include using an online grocery site or, for a more low-tech equivalent, a good old-fashioned shopping list.

The NHS Choices 5 A DAY meal planner can also help you plan your weekly shop.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

Hungry shoppers 'buy more calories'. BBC News, May 6 2013

Links To Science

Tal A, Wansink B. Fattening Fasting: Hungry Grocery Shoppers Buy More Calories, Not More Food. JAMA Internal Medicine. Published online May 6 2013

"Children who have overprotective parents are more likely to be bullied by their peers," BBC News explains.

The news correctly presents the findings of a major study on the effects of parenting on a child’s risk of being bullied, but it focuses on the weakest finding of the research.

The study did suggest that overprotective parents may increase a child’s risk of being bullied by their peers. However, the study also found that children with neglectful or abusive parents had an even greater increased risk of being bullied.

The headlines could also have focused on the more positive results – researchers found that children brought up in an emotionally warm environment with clearly defined rules about right and wrong were less likely to be bullied. This finding is interesting given the recent news about the potentially positive effects of ‘tough-love’ parenting.

Further studies into the association between parenting and a child’s chance of being bullied could shed additional light on the importance of a parent’s behaviour. While the findings of this study are interesting, it is not easy to see how it could be used to persuade people to change their parenting styles for the better.

Where did the story come from?

The study was carried out by researchers from the University of Warwick and Kingston University London and was funded by the Economic and Social Research Council and the Qatar National Research Fund.

The study was published in the peer-reviewed medical journal Child Abuse and Neglect.

Media coverage of this review largely focused on one of eight studied parenting styles (overprotection or “mollycoddling”). While the Daily Mail at least mentioned the larger detrimental effects of other parenting styles, some outlets (including the BBC News and the Daily Express) focused solely on the impact of overprotective parents.

The fact that the study found that a more positive parenting style – combining a mixture of emotional warmth and “firm but fair” rules – was linked with reduced chance of being bullied was not highlighted in the reporting of the study.

What kind of research was this?

The study was a combination of a systematic review and meta-analysis. It examined the relationship between parenting styles, parent-child relationships and bullying.

Researchers believe that family experiences and parenting style before children start school can influence the child’s capacity to adapt and cope at school. This can influence their relationships with schoolmates, making a child less, or more, vulnerable to bullying from their peers.

The researchers pooled the results from both prospective cohort studies and cross-sectional studies. This was to investigate the association between parenting behaviour and victimisation, and from this to identify parenting styles and family relationships that may increase the risk of victimisation.

What did the research involve?

The researchers searched the available literature for cohort and cross-sectional studies of the association between parenting behaviour and peer victimisation or bullying. They included studies that were published between 1970 and 2012 and provided a measure of relational, physical, verbal or cyberbullying.

The researchers identified parenting variables that they classified into positive and negative parenting behaviours.

The positive parenting behaviours were:

• authoritative parenting (highly demanding, but also highly responsive parents)
• parent–child communication
• parental involvement and support
• supervision
• warmth and affection

The negative parenting behaviours were:

• abuse or neglect
• maladaptive parenting (high levels of hostility, hitting and shouting)
• overprotection (or mollycoddling, as the media termed it)

The researchers included studies that recorded two types of child outcomes – victims and those who were both bullies and victims (bully/victims). They pooled the results of identified studies for each of these parenting styles to determine whether there were specific types of parenting behaviour that were associated with either the risk of being bullied or becoming a bully/victim.

The size of the effect of parenting on a child’s risk of being bullied or becoming a bully/victim was estimated using a statistical scale called 'Hedge’s g'. This scale is widely used to assess the impact of different types of effect or effect size. For example:

• a small effect would be a Hedge’s g measurement of 0.20
• a medium effect would be a Hedge’s g measurement of 0.50
• a large effect would be a Hedge’s g measurement of 0.80

A negative effect indicates a lower likelihood of victims of bullying having parents with that particular behaviour or style compared with non-victims.

What were the basic results?

The researchers identified 70 cohort and cross-sectional studies that met the inclusion criteria. These studies included 208,778 children and young people aged between 4 and 25 years. The researchers found that different parenting styles were associated with varying risk of being bullied or becoming a bully/victim.

Both victims and bully/victims were more likely to be exposed to negative parenting behaviour including abuse and neglect as well as maladaptive and overprotective parenting (effect size 0.26, 95% confidence interval (CI) 0.16 to 0.37). When examining the types of negative parenting styles, the researchers found that all had a significant association with victimisation, including:

• abusive or neglectful parents (effect size 0.31, 95% CI 0.18 to 0.44)
• maladaptive parenting (effect size 0.27, 95 CI 0.15 to 0.40)
• overprotective parents (effect size 0.10, 95% CI 0.03 to 0.17)

Conversely, overall, positive parenting behaviour had a small but significant effect, reducing the likelihood of the child being bullied or becoming a bully/victim (effect size -0.19, 95% CI -0.23 to -0.15). All five of their selected styles were associated with lower likelihood of being bullied:

• authoritative parents (effect size -0.19, 95% CI -0.28 to -0.11)
• good parent–child communication (effect size -0.12, 95% CI -0.20 to -0.05)
• involved and supportive parents (effect size -0.22, 95% CI -0.29 to -0.15)
• parents providing supervision (effect size -0.16, 95% CI -0.21 to -0.12)
• warm and affectionate parents (effect size -0.22, 95% CI -0.30 to -0.14)

For victims, the effects were generally small to moderate for positive parenting styles (effect size -0.12 to -22) and negative parenting styles (effect size 0.10 to 0.31). For bully/victims the effects were generally moderate for positive parenting styles (-0.17 to -0.42) and negative parenting styles (0.13 to 0.68).

How did the researchers interpret the results?

The researchers concluded that negative parenting styles are associated with “small to moderate effects on victim status at school” and that “intervention programs against bullying should extend their focus beyond schools to include families and start before children enter school”.

Conclusion

This research suggests that certain parenting styles may protect children against bullying risk. These include:

• being authoritative
• being involved and supportive
• being warm and affectionate
• having good communication with your child
• providing appropriate supervision

On the other hand, negative parenting styles were linked with an increased likelihood of being bullied. The researchers defined negative parenting styles as both “caring too much” or being overprotective and “not caring enough” or being neglectful.

Most of the headlines state that ‘mollycoddling’ your children increases their risk of being bullied. While these headlines are supported by this research, overprotective parenting styles were in fact associated with the smallest effect on bullying risk of the eight styles investigated.

The researchers point out that the other two negative parenting styles (abuse and neglect, and maladaptive parenting) were far more likely to increase the risk of a child being bullied.

The review assessed the effects of these parenting behaviours on the likelihood of the child both being a victim of bullying as well as bullying others. Generally, the relationships between parenting and the child bullying others were stronger than those between parenting and victimisation alone. Sadly, this more important finding was largely ignored by the media.

The researchers suggest that, “intervention programs that target children who are exposed to harsh or abusive parenting, may prevent peer victimization”. They also conclude that “parental training programs may be necessary to strengthen supportive involvement and warm and affectionate parenting to improve family relationships and prevent or reduce victimization by peers”.

Read more advice and information about bullying.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

Over-protected children 'more likely to be bullied'. BBC News, April 26 2013

Youngsters mollycoddled by their parents 'are more likely to be bullied by their peers'. Mail Online, April 26 2013

Spoiled children are most bullied. Daily Express, April 26 2013

Links To Science

Lereya ST, Samara M, Wolke D. Parenting behavior and the risk of becoming a victim and a bully/victim: A meta-analysis study. Child Abuse & Neglect. Published online April 25 2013

“Believing in God can help treat depression,” the Mail Online website claims. But how much faith can we put in this story?

The story is based on US research examining the association between belief in ‘God or a Higher Power’ and the effectiveness of mental health treatment.

The study found that patients with a self-reported strong belief in God were more likely to respond to treatment, and that a higher level of belief was associated with greater reduction in mental health symptoms such as depression and the desire to self-harm.

There are several important points to bear in mind when considering the results of this study. These include that:

• the type of study can only show an association, it cannot prove that belief in God will help people respond to treatment for depression
• the study was conducted in a small, specific population so it may not hold true for other groups of people
• belief in God was only measured by a single question, and the reliability and validity of this method is unclear
• it only examined religious belief and did not include the effect of secular (eg political) beliefs

Go to the Moodzone for more tips on changes you can make to your life if you are feeling down.

Where did the story come from?

The study was carried out by researchers from Harvard Medical School and was funded by the Gertrude B. Nielsen Charitable Trust – a US-based charity with a stated interest in childcare.

The study was published in the peer-reviewed Journal of Affective Disorders.

The Mail Online covered this story relatively well, but did not discuss the inherent limitations of the study. It also mentioned two additional studies, one apparently related to prayer and treatment of cardiac patients, and the other related to the success of IVF treatments. However, it failed to provide enough detail of these studies to allow us to assess what quality of evidence was on offer.

What kind of research was this?

This was a prospective cohort study that examined the association between belief in ‘God or a Higher Power’ with outcomes for patients being treated for mental health disorders.

The researchers report that previous studies suggest that spiritual or religious beliefs may act as a buffer against several mental health conditions and behaviours, including depression and self-harming.

Some studies suggest, however, that spiritual struggles can worsen or bring on symptoms.

As a cohort study, this research cannot tell us about any potential causal link between belief and treatment, only whether the two factors are associated. Furthermore, it cannot tell us what it is about belief that leads to an association with treatment outcomes.

What did the research involve?

The researchers recruited 159 patients in a day-treatment programme at a psychiatric hospital in the US. The average age of the patients was 34 years, approximately 62% of whom were women. All patients were experiencing serious symptoms or impairment. Mental health disorder diagnosis varied across the participants, with 60% having major depression, 12% bipolar disorder, and the remaining 28% having various other diagnoses including anxiety.

Before treatment, the researchers measured the patients’ belief in God by asking a single question: “to what extent do you believe in God?”, measured on a five-point scale from “not at all (no belief at all)” to “very (a strong sense of belief)”.

The researchers followed up the patients over the course of a year, and assessed four main treatment outcomes:

• treatment response
• degree of reduction in depression symptoms over the course of treatment 
• overall psychological wellbeing
• self-harm behaviours

During the analysis, the researchers controlled for both age and gender as potential confounders, as both were associated with religious belief. They also assessed a range of variables they thought may account for, or mediate, any relationship between belief and treatment outcomes.

These factors included:

• their beliefs about treatment, including credibility (how confident the patients would be recommending the treatment to a friend who was having the same problems), and treatment expectancy (how much improvement in symptoms the patients expected to experience by the end of treatment)
• emotion regulation, which included an assessment of both positive and negative strategies to control emotions 
• degree of support provided by the patients’ congregations, based on two questions regarding the extent to which the patients obtained emotional support from the spiritual or religious communities

What were the basic results?

The researchers found that belief in God or a higher power was significantly higher among those patients who responded to treatment compared to those that did not. Also, a higher level of belief was linked with greater reduction in depression symptoms and self-harm behaviours, and greater gains in overall psychological wellbeing over the course of treatment.

The type of religious affiliation – such as Catholic, Jewish or Hindu – had no effect on treatment response or any other psychological or behavioural variables.

Belief in God remained significantly associated with changes in depression and self-harm even after controlling for the patients’ age and gender, two factors that could potentially confound the relationships. The patients’ perceptions regarding treatment credibility and expectations about treatment effects were associated with belief in God.

None of the other variables looked at by the researchers were found to significantly change the relationship between belief and self-harm or psychological wellbeing.

How did the researchers interpret the results?

The researchers concluded that, “belief in God, but not religious affiliation, was associated with better treatment outcomes. With respect to depression, this relationship was mediated by belief in the credibility of treatment and expectations for treatment gains”.

Conclusion

This research suggests that religious or spiritual belief may be associated with response to treatment for some mental health disorders. However, the study cannot tell us what aspects of belief may be important in terms of this link to depression symptoms, treatment response and overall psychological wellbeing.

The researchers say that their findings suggest that “belief in the credibility of psychiatric treatment and increased expectations to gain from treatment might be mechanisms by which belief in God can impact treatment outcomes”.

They say that it is “notable that faith in treatment was virtually not present in the absence of belief in God, and that few participants with high belief in God had low treatment credibility/expectancy”. They also say that “this may suggest that faith is a general cognitive attribute” that may represent an optimistic outlook in several areas, including the spiritual and medical.

There are some limitations to the study that should be considered, including the facts that:

• All of the participants in this study were in a day-treatment programme for mental health conditions, and all were experiencing symptoms that severely limited their functioning. These participant characteristics make generalisation to less severe forms of these disorders difficult. It is also important to note that the majority of people in the study being treated for mental health conditions (61.6%) reported having a belief in God or a higher power.
• The positive effects of secular or political beliefs were not studied by the researchers.
• The study was highly culturally specific: the majority of participants expressing a religious belief were Christian.
• Belief in God was assessed using a single question, with no mention of the reliability or validity of this question in measuring belief.

This study provides insight into the relationship between faith, or belief, and mental health, and suggests a potential path through which such an association may operate.

Further research could be carried out to measure the size of the effects that believing in a ‘higher power’ (whether a supreme being or a concept of ‘humanity’ and ‘goodness’) may have on mental health outcomes.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

The power of prayer: Believing in God can help treat depression. Mail Online, April 25 2013

Links To Science

Rosmarin DH, Bigda-Peyton JS, Kertz SJ, et al. A test of faith in God and treatment: The relationship of belief in God to psychiatric treatment outcomes. Journal of Affective Disorders. Published online October 7 2013

"TV chefs 'adding to obesity crisis by encouraging us to eat fatty dishes'," reports the Metro, with similar stories blaming celebrity chefs for our bulging waistlines in much of the media.

The news is based on analysis of the nutritional values of randomly selected recipes created by celebrity chefs. Researchers found that most of the recipes analysed exceeded national healthy eating benchmarks on fat, saturated fat, sugar and salt intake.

The problem with this study, and the media’s reporting of it, is that it assumes that unhealthy recipes lead to higher rates of obesity, which has not been shown to be the case.

We cannot draw reliable conclusions on the effects of these findings because, for example, we don’t know if these recipes are cooked and eaten frequently and we don’t know how other recipes compare.

It seems unlikely that cooking some of these recipes for a special occasion or as a treat will harm your health, especially if you eat a balanced diet and compensate for treats with healthier options at other mealtimes.

Read more about healthy eating.

Where did the story come from?

The study was carried out by researchers from the Department of Health Professions at Coventry University, and was supported by the department and Faculty of Health and Life Sciences, Coventry University.

The study was published in the peer-reviewed journal Food and Public Health and is freely available on an open-access basis.

This story was widely covered, with nearly all papers leading with a headline blaming TV chefs for making us fatter or adding to the obesity crisis. The researchers’ claims that “Celebrity chefs are a likely hidden contributing factor to Britain’s obesity epidemic” were accepted without any serious scrutiny. The study presents no credible evidence that cookbooks containing unhealthy recipes are directly responsible for obesity rates.

Celebrity chefs can have a positive effect on people’s diets, and to blame them for Britain’s rising obesity levels is an oversimplification of the problem.

Somewhat unfairly, many of the news stories featured a photo of Nigella Lawson. While some of her recipes may be unhealthy, the celebrity chefs included in this study were not named.

A similar study, published in 2012, comparing ready meals with celebrities’ recipes was reported in the same uncritical way.

What kind of research was this?

This was a cross-sectional study that analysed the nutritional composition of British-based celebrity chefs’ recipes.

The researchers wanted to compare the nutrient content with national benchmark recommendations for certain nutrients and healthy eating guidelines, such as the advice to eat less than six grams of salt a day.

Although this study allows conclusions about the nutritional content of recipes to be drawn, it does not assess the impact of celebrity chefs’ recipes on people’s diets. The findings of this study do not allow us to determine whether celebrity chefs are “adding to the obesity crisis”, because, for example, we don’t know how often these meals are eaten.

What did the research involve?

Recipes by celebrity chefs were randomly selected. Celebrity chefs were professional cooks whose recipe books appeared on Amazon’s “top one hundred bestselling books of 2009” or who were featured on the Good Food Channel’s website as a celebrity chef.

To be eligible, recipes had to be suitable for the general public, rather than being targeted at, for example, children or people who wanted to lose weight.

In total, 904 recipes from 26 celebrity chefs were selected at random. All types of recipes (for example breakfast, lunch, starter, evening meal and dessert) were eligible for inclusion.

Researchers used computerised dietary analysis software to work out nutritional content of the recipes. For each recipe they worked out the levels of:

• total energy
• protein
• carbohydrate
• fat
• sodium
• salt

The nutritional value of each recipe was compared against national healthy eating benchmark guidelines using a ‘healthy eating index’, which measured how far each recipe deviated from national recommendations.

What were the basic results?

Recipes from the 26 chefs differed significantly in the energy, protein, carbohydrate, fat and salt and sodium content per portion.

Per portion:

• Recipes from 22 of the 26 chefs had fat content on average above the ‘high fat content’ criteria.
• Recipes from 24 of the 26 chefs had saturated fat content on average above the ‘high saturated fat content’ criteria.
• Recipes from 16 of the 26 chefs had sugar content on average above the ‘high sugar content’ criteria.
• Recipes from seven of the 26 chefs had salt content on average above the ‘high salt content’ criteria.

On average, recipes had nutritional levels substantially above healthy eating guidelines (meaning that on average they contained more fat, saturated fat, sugar and salt than recommended).

The researchers calculated that 13% of recipes met or were below healthy eating guidelines (meaning that on average they had less fat, saturated fat, sugar and salt than the recommended limits), whereas 87% exceeded healthy eating guidelines.

Perhaps unsurprisingly, when the meals were analysed by meal type subgroup, desserts were the meal type that exceeded healthy eating guidelines the most.

How did the researchers interpret the results?

The researchers conclude that, “although variation in the nutritional composition of recipes existed between celebrity chefs, there was still a general trend whereby excessive amounts of total fat, [saturated fat], sugars and salt were evident. The majority of recipes analysed had unhealthy nutritional compositions in accordance with national healthy eating benchmark recommendations, and therefore celebrity chefs could potentially be a hidden contributory factor to current public health nutrition issues, through exacerbating Britain’s already unbalanced dietary intake”.

Conclusion

This study has found that many celebrity chefs’ recipes exceed national recommendations for fat, saturated fat, sugar or salt intake.

However, although the researchers and the media have speculated on the effect that this may have, this research does not investigate this question and no conclusions can be drawn. For example, we don’t know if these recipes are cooked and eaten frequently, and we don’t know how the nutritional value of these celebrity chefs’ recipes compares with more humble cooks’ recipes.

It is also important to repeat the fact that celebrity chefs who targeted their recipes at people concerned about weight management or who were on a diet were excluded from the study.

Often, TV chefs’ recipes are designed to be ‘event meals’, with the meal being cooked for a special occasion such as a birthday or dinner party. It is unlikely that someone would use a cookbook to cook all their meals.

It is similarly unlikely that cooking some of these recipes for a special occasion or a treat will harm your health, especially if you eat a healthy balanced diet the rest of the time.

Read more about healthy eating and good food.

Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.

Links To The Headlines

TV chefs ‘adding to obesity crisis by encouraging us to eat fatty dishes’. Metro, April 24 2013

Television chefs adding to obesity crisis with fatty dishes warn academics. The Independent, April 24 2013

Celebrity chefs 'exacerbate Britain's obesity crisis by encouraging people to eat dishes packed with fat'. Daily Mail, April 24 2013

TV chefs get roasting for 'making us fatter'. Daily Express, April 24 2013

Links To Science

Jones M, Freeth EC, Hennessy-Priest K, Costa RJS. A Systematic Cross-Sectional Analysis of British Based Celebrity Chefs’ Recipes: Is There Cause for Public Health Concern? Food and Public Health. Published online 2013