The obesity paradox in older age groups

Levels of obesity in the population increase with age until around 70 years, after which prevalence declines due to a number of factors including concurrent illness, such as chronic obstructive pulmonary disease (COPD) and malignant disease causing weight loss; failure to thrive of the elderly; and the fact that obese individuals tend to die prematurely. However, with improving healthcare and public health, life expectancy is increasing and it can be anticipated that the peak age for obesity will increase as will both the clinical and economic burden for an extra decade.

Further reading: 

• Obesity-related health inequality could overwhelm NHS, new reports says
• New government services launched to tackle rising obesity rates
• Focusing on emotions improves the physical and mental health of people living with obesity

The weight of men peaks earlier than women, possibly due to lower life expectancy, and deterioration in health at an earlier age.⁴ In contrast, when Body Mass Index (BMI) is plotted and the prevalence of obesity is diagnosed by waist circumference, a different picture is revealed that shows increases are unabated with age, probably demonstrating the vagaries of the different measuring techniques. The higher rate of obesity as defined by waist circumference probably represents those individuals displaying sarcopenia, with loss of muscle mass, in favour of an expanded visceral adipocyte mass.

To further complicate the picture, height reduction in the older patient due to changes in bone and disc composition affects the interpretation of measurements. If a subject decreases in stature due to loss in vertebral bone, there is a corresponding increase in BMI.⁵

HSE data also reveal increasing levels of obesity in immigrant and deprived populations, demonstrating the health inequalities of obesity, and raising the possibility that, as both subsections of the population are increasing, the obesity epidemic will worsen imminently.

The Foresight Report, published in 2007 by the Department of Trade and Industry suggests that by 2050, 60% of men and 50% of women could be clinically obese, and that without action, obesity-related diseases will cost an extra £45.5 billion per year to the economy.⁶

A report by McKinsey⁷ also revealed that obesity has a similar Global Economic Impact as armed violence, war and terrorism combined. In response, over recent months, major figures in Government, Public Health England and the NHS have started to become serious about obesity and its ramifications.

Simon Stevens, Chief Executive of the NHS, launched the Diabetes Prevention Programme across the country as a means of tackling obesity, and prioritised obesity in his Five Year Plan.

Chief Medical Officer Professor Dame Sally Davies said: €˜We must get to grips with the problem now to save lives and money in the future.’ In addition, the Government Health Select Committee has promoted obesity treatments and the Royal College of Physicians, the Royal College of GPs and the Academy of Royal Colleges have all recently set up formal stakeholder groups to tackle the problem. The newly created Obesity Health Alliance collaboration is also set to enhance anti-obesity policies and activities.

Clinical aspects of obesity

Rather than being merely a cosmetic or social issue, the seriousness of an expanded adipocyte mass lies in the profound deleterious effect it has on all the systems of the body. This is by its mechanical presence, hormonal and psychological ramifications, but mainly in its metabolic influence centred on the production of pro-inflammatory cytokines, reduced adiponectin, and the induction of insulin resistance. Excess body weight in the elderly correlates strongly with chronic ill health, poor quality of life, functional decline, disability, and dependency.⁸

The most widely accepted comorbidity of obesity is type 2 diabetes and approximately 20% of the population develop diabetes by the age of 75.⁹ Over the age of 65 years, diabetes or glucose intolerance was present in 30€“40% of Framingham Study subjects.¹⁰ The fact that half of subjects are unaware that they have the condition¹¹ emphasises the importance of screening for diabetes in the elderly obese population.

The improvement of identification and management of diabetes, especially over the last decade, paradoxically, is being offset by enhanced life expectancy, and therefore increased likelihood of long-term microvascular, microvascular and other complications. These include the more traditionally recognised conditions such as cardiovascular disease, retinopathy, nephropathy and neuropathy, as well as less readily associated ones such as cardiac autonomic neuropathy,¹² Charcot neuroarthropathy,¹³ periodontal disease¹⁴ and falls.¹⁵ Elevated BMI is not only associated with a diagnosis of diabetes, but is also linked with progressively higher risk for all diabetes complications, especially in women. 

One study showed that for women, being slightly overweight led to higher risk of cardiovascular, cerebrovascular, renal, and lower extremity complications. For men, slightly overweight status was not significant for developing diabetes complications. For women in the upper range of overweight, the risk also increased for cardiovascular, renal, ocular, and lower extremity complications. Men too achieved greater risk compared to men with normal BMI.¹⁶

Rates of cardiovascular disease are increased in association with central obesity,¹⁷ independent of diabetes, partially due to the pro-inflammatory environment, including endothelial dysfunction and blood hypercoagulability, promoted by adipocytokines produced by visceral adipocytes such as TNFα and PAI-1.¹⁸

Obstructive sleep apnoea is also associated with obesity and much more prevalent in the elderly.¹⁹ However, there are comorbidities which, although not unique to the elderly, have specific ramifications for this group. Chronic venous insufficiency,²⁰ for example, is linked with obesity; in one study 45.2% of leg ulcer sufferers were obese, whilst only 27.2% had non-insulin dependent diabetes.²¹ Varicose veins are also more common in obese subjects.²² 

Framingham demonstrated links between obesity and osteoarthritis, stating €˜cross-sectional data show that obesity or as yet unknown factors associated with obesity cause knee osteoarthritis.’²³ There also seems to be a connection between obesity and inflammatory arthritis.²⁴ 

Findings from NHANES²⁵ suggest that obesity is associated with depression mainly among persons with severe obesity. Functional aspects of obesity include breathless on daily activities and climbing the stairs, and inability to manage simple personal hygiene tasks, and cutting toenails.

Sarcopenia and obesity in older age

For the diagnosis of sarcopenia, the European Working Group on Sarcopenia in Older People (EWGSOP)²⁸ recommends using the presence of both low muscle mass and low muscle function (strength or performance). Sarcopenia with limited mobility is defined as a person with muscle loss whose walking speed is equal to or less than 1m/s or who walks less than 400m during a six-minute walk. Also who has a lean appendicular mass corrected for height squared of 2 standard deviations or more below the mean of healthy persons between 20 and 30 years of age of the same ethnic group.

The limitation in mobility should not clearly be a result of otherwise defined specific diseases of muscle, peripheral vascular disease with intermittent claudication, central and peripheral nervous system disorders, or cachexia. Clinically significant interventions are defined as an increase in the six-minute walk of at least 50 meters or an increase of walking speed of at least 0.1 m/s.²⁹

When obesity and sarcopenia coincide, the result is sarcopenic obesity;³⁰ at any given BMI, an older individual will have a higher percentage of body fat at the expense of muscle, than a younger person. Increased physical activity in the elderly can produce beneficial effects on muscle strength, endurance, and well-being³¹ and is therefore a vital component of weight management in this age group. The expression €˜dynapenic obesity’ describes the combined effect of obesity and low muscle strength, which has the effect of increasing the risk of incident type 2 diabetes in older adults.³²

A 10% weight loss is associated with clinically significant reductions in hypertension, lipids mortality,³³ sleep apnoea³⁴ and other comorbid conditions. The Look-AHEAD study³⁵ intended to answer the ultimate question: does weight loss by lifestyle changes reduce cardiovascular mortality? The study looked at 5,000 patients with type 2 diabetes for 13.5 years following intensive initial treatment, then less intensive subsequent input. Four-year data revealed impressive weight loss, and improvement in surrogate markers, in particular, a 23% reduction in insulin users from baseline. Remarkably, despite these outstanding results, Look-AHEAD was halted for reasons of €˜futility’ due to a lower than expected event rate. 

In 2012, a 13-year follow up paper to the landmark Diabetes Prevention Study was published. The original paper famously demonstrated a 58% reduction in cumulative incidence of diabetes in subjects with impaired glucose tolerance, with intensive lifestyle intervention. The follow-up study demonstrated a €˜legacy’ effect€”those people who initially underwent intensive lifestyle treatment benefited from a long-term 32% reduction in the cumulative incidence of diabetes, despite lapses in diet and physical activity after the premature culmination of the trial.³⁶

The obesity paradox

Clearly, obesity is associated with comorbidities including diabetes, cardiovascular disease, obstructive sleep apnoea, cancer, and many others in the elderly population. Equally clearly, weight loss confers protection against their onset and may improve these conditions. It is counter-intuitive, therefore, that, although obesity is implicated in their cause, its presence seems to be protective against adverse outcomes once they have occurred. €œThe idea that a known risk factor somehow transforms into a €˜protective’ agent after an occurrence of a vascular clinical event is both surreal and troubling€.³⁷ This phenomenon is known as the €˜Obesity Paradox’; an example of so-called €˜reverse epidemiology’.³⁸

There is an increasing body of data contradicting traditional weight loss assumptions in elderly individuals, suggesting that although excess weight clearly contributes to various comorbid conditions, once those conditions have occurred, excess weight may improve subsequent outcomes.³⁹ Evidence from studies including 30,000 chronic heart failure patients suggests that overweight and obese patients with heart failure had reductions in CV (€“19% and €“40%, respectively) and all-cause (€“16% and €“33%, respectively) mortality during 2.7-year follow-up.⁴⁰ A similar pattern occurs in acutely decompensated heart failure; higher BMI is associated with lower in-hospital mortality.

A meta-analysis of 250,000 patients with coronary artery disease, cardiovascular and total mortality outcomes were better in overweight and €˜mildly’ obese patients compared with €˜normal’ weight.⁴¹ The INVEST study of 22,500 individuals with hypertension plus coronary artery disease, showed a lower risk of death or major cardiovascular events in overweight and obese subjects compared with normal weight.⁴² The Scottish Coronary Revascularisation Register concurred:⁴³ among patients undergoing PCI for CAD, raised BMI was associated with enhanced 5-year survival. 

Another study showed that overweight and obesity are associated with improved short- and long-term survival after acute myocardial infarction, resulting in moderate gains in life expectancy relative to normal-weight patients.⁴⁴ These findings suggest that higher BMI confers a protective advantage over the entire remaining lifespan in older patients with acute myocardial infarction. Other publications have shown that higher percentage body fat predicts better prognoses in HF patients, and possibly CHD mortality.⁴⁵ Meta-analysis has also shown that a BMI of 33.4-60€”too high for there to be any doubt of excess adiposity€”there is significantly higher LV ejection fraction, with significant BNP decreases in the highest BMI quartile.⁴⁶

Some studies specifically on body fat assessed by dexa-scan show that with established CV diseases, including advanced HF, higher levels of BMI and Body Fat appear to be protective,⁴⁷ except in extreme obesity (BMI) >40),48 however there is emerging contradictory evidence that benefits might only be short-lived.⁴⁹ Subjects with peripheral vascular disease benefit in a similar fashion; mortality decreases with increasing BMI, possibly explained by the presence of COPD inducing weight loss.⁵⁰ An unexpected survival benefit has been reported in sleep apnoea associated with obesity.⁵¹

Obese and overweight CVA subjects have significantly better survival than their counterparts;⁵² obesity is associated with a lower stroke and mortality rate in elderly anticoagulated atrial fibrillation patients.⁵³ Obesity is also linked with lower mortality in pre-capillary and disproportional post-capillary pulmonary hypertension patients.⁵⁴ In addition, it is a risk factor for end stage renal failure, but its presence may improve outcomes.⁵⁵

Studies of dementia, including Alzheimer’s disease, demonstrate a similar pattern: obesity in middle age is related to greater dementia risk in later life,⁵⁶ but when later life materialises, the risks are reversed.⁵⁷ Less controversially, low BMI is associated with an increased fracture risk, whereas a high BMI is protective in all ages, and with all, but mainly hip fracture.⁵⁸

Early research suggests that higher BMI may have a protective effect against mortality in vascular patients with lower limb ulcers.⁵⁹ To add to the uncertainty, a recent JAMA paper suggests that individuals diagnosed with type 2 diabetes with relatively low BMIs fared less well in terms of mortality than overweight and obese people.

However, the obvious reaction that obesity might be somehow protective against diabetes, needs scrutiny, being entirely counter-intuitive. Possible explanations include the use of BMI rather than direct measures of adiposity being misrepresentative, or that patients carrying excess weight are identified younger, enabling enhanced prevention by cholesterol and blood pressure optimisation; low to normal weight may be caused by intercurrent disease, failure to thrive of the elderly, excess alcohol intake or smoking, and weight loss may be unintentional. Furthermore, BMI may be high because of athletic build and muscularity€”protective against CVD€”in contrast to increased adiposity, known to be harmful.

Excess adiposity may act as a protective €˜metabolic reserve’; some individuals may only suffer from a condition because of their obesity, and therefore suffer a less malign form of the illness than someone who receives the diagnosis despite being lean.

For example, type 2 diabetes is unusual in lean individuals, so might represent a more sinister manifestation of the disease. People may have been diagnosed later in the course of the illness because of their leanness, or may have lost weight because of previously undiagnosed diabetes. However, the number of studies supporting the obesity paradox is growing and possible confounders identified and accounted for, but evidence is still strong.

Dr David Haslam is a GP and Physician specialising in Obesity Medicine at the Centre for Obesity research at Luton & Dunstable Hospital and fomer Chair of the National Obesity Forum. 

Conflict of interest; none declared