Mental health

Eating disorders in middle-aged women 'common'

"Eating disorders…affect a small but substantial number of women in their 40s and 50s," BBC News reports. While often regarded as a "disease of the young", a new survey suggests 3.6% of middle-aged women in the UK are affected by an eating disorder.

Researchers also looked at childhood, parenting and personality risk factors associated with the condition. They found that 15% of middle-aged women had experienced an eating disorder at some point in their lifetime, and 3.6% had one in the last 12 months.

A commonly reported disorder is what is known as "other specified feeding and eating disorder". This term describes cases where a person may not fit the precise pattern of eating disorders such as anorexia but they still experience significant distress due to an unhealthy psychological relationship with food.

The study found that all potentially harmful childhood life events such as child sexual abuse, death of a carer and parental divorce, were associated with the onset of eating disorders. However, the study can't prove that these factors caused the disorder.

The researchers hope this survey will highlight that when it comes to diagnosing eating disorders, health service provision for middle aged women could be improved.

Read more about the help available for people with eating disorders as well as advice for friends and family who may be worried about others.

Where did the story come from?

The study was carried out by researchers from several UK, US and Swedish institutions including University College London, Harvard Medical School and the Karolinska Institutet in Stockholm. It was funded by the National Institute of Health Research UK and the UK children's charity Wellchild.

The study was published in the peer-reviewed medical journal BMC Medicine on an open-access basis, so it is free to read online.

BBC News provided a well-balanced report on the study.

In contrast, the Daily Mail's reporting was both confused and confusing. Its headline: "Divorce blamed as more middle-aged women are hit by eating disorders", would naturally lead readers to assume it is going through divorce is a risk factor. But the study only explicitly mentions parental divorce as a risk factor in childhood.

What kind of research was this?

This was a cross-sectional analysis which used data from an existing longitudinal study – the UK Avon Longitudinal Study of Parents and Children (ALSPAC) to investigate the prevalence of eating disorders in middle-aged women. Within this, the researchers explored childhood, parenting and personality risk factors associated with the condition.

Eating disorders are severe mental health problems which cause an individual to change their eating habits and behaviour. The conditions can affect someone physically, psychologically and socially.

Traditionally, these conditions are associated with younger women but the researchers recently identified a gap in access to healthcare for adults with eating disorders in a UK population. As a result, they wanted to investigate this further.

Observational studies like this one are useful for assessing the incidence and prevalence of health conditions. However the study design limits the ability to prove causation between exposure and outcome, for example, between a potential risk factor and the development of an eating disorder. 

What did the research involve?

The data for this analysis was obtained from ALSPAC, a population-based prospective cohort study of women and their children.  ALSPAC followed 14,541 pregnant women and examined the effects of environment, genetic and other factors on them and their children.

This analysis included a sample of 9,233 of the women (average age 48 years) and asked them to complete a version of the Eating Disorders Diagnostic Schedule (EDDS). The EDDS uses different criteria to diagnose the following conditions:

Women who were screened positive (5,655) based on the screening criteria were interviewed using the eating disorders section of the Structured Clinical Interview for DSM-IV-TR disorders (SCID-1).

The interview assessed presence, frequency and duration of behaviours associated with eating disorders such as restriction, fasting, excessive exercise, binge eating, and purging. The women were asked to relate any changes in their eating behaviours with major life events to see if they were potentially associated.

Data for 1,043 women on relevant predictors of the onset of eating disorders was obtained from the ALSPAC database collected 20 years prior to this analysis:

  • childhood unhappiness
  • parental divorce or separation, adoption or being under health authority care
  • death of a carer
  • early sexual abuse
  • life events
  • bonding with parents
  • locus of control (LOC ) – whether a person feels in control of their life
  • interpersonal sensitivity

The data was then analysed to search for any potential associations between risk factors and the onset of eating disorders.

Potential confounders such as maternal age, ethnicity and education were adjusted for.

What were the basic results?

Overall the researchers found 15% of middle aged women had experienced an eating disorder in their lifetime, and 3.6% had one in the last 12 months.

Anorexia nervosa was the most common specific lifetime disorder, with a prevalence of 3.6%, though the general category of "other specified feeding and eating disorder" was most common, affecting 7.6%.

Several links emerged between early risk factors and the onset of eating disorders:

  • Experiencing the death of a carer was associated with a seven-fold increase in odds for the onset of purging disorder (odds ratio [OR] 7.12; 95% confidence interval [CI] 2.32 to 21.85).
  • There were higher odds of suffering from bulimia nervosa (OR 2.02), binge eating disorder (OR 2.01) and anorexia nervosa (OR 2.49) following parental separation or divorce in childhood.
  • Child sexual abuse was associated with all disorders linked to binge eating behaviours: anorexia nervosa binge purge (OR 3.81), bulimia nervosa (OR 4.70) and binge eating disorder (OR 3.42).
    Sexual abuse from a non-stranger was linked with anorexia nervosa binge purge, bulimia nervosa and binge eating disorder.
  • Childhood unhappiness was associated with increased odds of anorexia nervosa (OR 2.52), bulimia nervosa (OR 4.58), binge eating disorder (OR 3.66) and purging disorder (OR 2.65).

Overall, all the childhood life events were positively associated with eating disorders, and the more life events there were, the higher the risk.

How did the researchers interpret the results?

The researchers concluded: "Although some risk factors differed across [eating disorder] subtypes, childhood sexual abuse and poor parenting were associated with binge/purge type disorders, whilst personality factors were more broadly associated with several diagnostic categories. Few risk factors were specifically associated with one diagnostic category."


This well-designed cross-sectional analysis used data from an existing longitudinal study to investigate the prevalence of eating disorders in middle-aged women and see what childhood, parenting and personality risk factors were associated with the onset of an eating disorder.

The research found that more than 1 in 10 middle aged women experience some form of eating disorder in their lifetime. It found that all potentially harmful childhood life events such as child sexual abuse, death of a carer and parental divorce, were associated with the onset of eating disorders.

An association with traumatic life events is definitely plausible, or even likely. However, it must be noted that within the context of observational survey data, such studies are never able to prove that any single exposure causes the development of an eating disorder.

This study has not been able to take into account all aspects of a person's mental and physical health, interpersonal relationships and lifestyle prior to the onset of an eating disorder. Therefore the study can show associations but cannot prove definite causation with any individual factor.

The researchers say that this research has implications for health service provision in the UK, which needs to recognise that women in mid-life can still be suffering from the effects of long-standing disorders, or be at risk of developing new disorders. Therefore better awareness of eating disorders and their symptoms is needed.

Dr. Agnes Ayton, Vice Chair of the Faculty of Eating Disorders, Royal College of Psychiatrists commented on the research saying:

"This is an important paper, which has several methodological strengths: it is population-based (rather than only including people who seek contact with health care, which is always the tip of the iceberg). It has used reliable assessment of the eating disorder, by interviewing with validated instruments, rather than relying on self-report. It was also able to identify risk factors, which were collected many years ago as part of the AVON Longitudinal Study, therefore avoiding recall bias.

"It demonstrates that the rates of eating disorders amongst middle age women are higher than it was thought, and that significant proportions of these people are unknown to services – so there is a large unmet need."

Find eating disorders support services in your local area.

NHS Attribution