This site is intended for health professionals only

At the heart of general practice since 1960

10 top tips

Low cholesterol may be just as much cause for concern as high cholesterol in the elderly population, conclude Bandolier editors Dr Andrew Moore and Professor Henry McQuay

As doctors we spend most of our time lowering cholesterol levels, but when it comes to our elderly patients the picture is not so clear. As we found in this review, cholesterol levels frequently fall in later years, especially in people over the age of 80 and who had high levels previously.

This may seem surprising, but as you will see, falling levels are frequently a marker for ill-health and functional decline generally, so it seems less surprising that low serum cholesterol should be associated with increased all-cause mortality in the elderly.

Of course there are many confounding factors in this age group, such as other medical conditions and multiple medications. Both tend to increase with age, while kidneys, liver, lungs, heart and muscle become less efficient. Older people tend to eat less as well.

To help clarify the picture, we reviewed five recent international studies with the hope of finding some common threads and providing a more integrated view. The first is the Honolulu study, which clearly shows an increased mortality associated with low serum cholesterol.

1 Mortality/low cholesterol ­ Honolulu1

This was part of a large epidemiological study starting in 1965 with 8,000 Japanese/ American men born between 1900/19. The study was based on examinations in 1991/3, with mortality to the end of 1996. The examination consisted of demographics, function tests, blood tests, and ECG.


There were 3,572 men aged 71-93. Mean serum cholesterol fell from 5.0mmol/L in those aged 71-74, to 4.6 in those older than 85. Analysis by quartiles of serum cholesterol showed that age-adjusted mortality was highest in the quartile with the lowest serum cholesterol.

Men in the lowest quartile for total cholesterol were more likely to have a history of weight loss and poor physical functioning. Much of the statistical relationship between low cholesterol and increased mortality was lost when risk factors and frailty measures were taken into account, but it remained in men who had low total cholesterol for about 20 years.

2 Mortality/cholesterol level ­ Italy2

In 1985 3,282 Italians aged 65 or older were screened, providing demographic data, blood samples, and cardiovascular and respiratory measurements. Subsequent mortality was assessed annually by monitoring discharge records and death certificates.


There were slightly more women than men, the mean age was 74 and the mean total cholesterol 5.7mmol/L. The average BMI was 26, and 61 per cent were classified as overweight. More men smoked, and more women had diabetes diagnosed. Total follow-up was 12 years, with 1,599 deaths.

For women, but not men, overall mortality was significantly higher in those with the lowest cholesterol (below 5.1mmol/L). Both men and women in the lowest quartile of cholesterol (below 5.1 for women and 4.7 for men) had increased mortality compared with participants with higher cholesterol and BMI above 25.

Women and men in the lowest quintile of cholesterol had higher rates of death from cancer or other causes than those in the highest quintile. For men but not women, the highest quintile of cholesterol conferred an increased risk of coronary heart disease death.

3 Cholesterol levels/age ­ Italy3

A second study from Italy included patients aged 65 or older admitted to 35 clinical centres (geriatric and internal medicine wards) during May-June and September- October 1995 with data on demography, diagnosis, and blood and other analyses.


There were 2,486 patients admitted (53 per cent women), with an average age of 78, and average total cholesterol of 4.7mmol/L.

Total cholesterol in men aged over 85 was 12 per cent lower than average than in those aged 65-69. Total cholesterol in women aged over 85 was 6 per cent lower than average than in those aged 65-69.

Disability indices were higher in those with lower cholesterol, as was the incidence of low serum albumin and iron, and a BMI below 20. Low cholesterol was independently associated with a diagnosis of infectious disease, low serum albumin and low BMI. After adjustment for these factors, the likelihood of having low serum cholesterol did not increase with age.

4 Cholesterol predicting functional decline ­ Holland4

A longitudinal study in a random sample of Dutch adults aged 55 to 85 included data on serum albumin and total cholesterol at baseline. They were asked at the beginning and three-year follow-up to perform three simple actions: walk three metres, turn around, and walk three metres back; stand up five times from a kitchen chair with arms folded across their chest; and put on and take off a cardigan. Each was timed, and the times scored according to a simple system.


The mean age of 1,064 participants was 68. Low serum albumin (below 43g/L) was found in 24 per cent of women and 22 per cent of men. Low serum cholesterol (below 5.2mmol/L) was found in 8 per cent of women and 18 per cent of men. A three-year decline in functional performance was associated with low cholesterol in women. In men, functional decline was associated with both low cholesterol and low albumin.

5 Statin use in older people ­ Ontario5

The geriatric Ontario longitudinal database linked several health care databases with follow-up of mortality over time. It includes 1.4 million residents of Ontario who were 66 or older in 1998. Information was obtained on medicines prescribed in the year before the cohort began.

The cohort comprised those at high risk of future cardiovascular events, with a history of cardiovascular disease or diabetes. Patients with recent history of cancer were excluded. The final cohort included 396,000 people.

A baseline risk index was stratified at 25 per cent, 50 per cent and 75 per cent, percentiles of death. Low, intermediate and high-risk patients were identified.


Two-thirds of the cohort had cardiovascular disease, 18 per cent diabetes, and the remainder had both conditions. Half were women, and the average age was 75. Statins were prescribed in 19 per cent. Patients prescribed statins were younger, more likely to be men, with a prior heart attack, angina, or invasive procedure, and made more visits to a cardiologist.

Patients not prescribed statins were more likely to have diabetes, congestive heart failure, stroke, or live in rural areas.

Patients who were older, or at higher baseline risk of dying over the next three years, were less likely to have a statin prescribed. Progressively lower use of statins in patients with higher cardiovascular risk existed across the full spectrum of risk, and across the entire spectrum of age.

What these studies tell us

These studies show that average cholesterol levels fall with increasing age, and low cholesterol is associated with increased mortality, particularly cancer and causes of death other than cardiovascular, which remains a major cause of death in older people.

Cholesterol loss appears to be greater in those with an initially high level. Low cholesterol is associated with low weight or BMI, or weight loss, poor physical functioning, infections, and other markers of ill-health, like low serum albumin and iron. Reduced hepatic synthesis, appetite and food intake are expected with increasing age.

Superimposed might be differing patterns of co-morbidity, and the propensity of some people to age faster than others.

When it comes to statin prescribing for older patients, many factors have to be considered. Older people at very high risk probably have additional health problems, possibly of more immediate importance than future cardiovascular events. Prescribing a statin is likely to be low on the list of priorities, particularly where many drugs are already prescribed.

We might also consider low cholesterol and albumin levels as markers of poor health. The difficulty is knowing what the cut-off should be for low. It might be 4.0mmol/L or below in people not on statins, but lower in those on statins or with previous cardiovascular disease.


1 IJ Schatz et al. Cholesterol and all-cause mortality in elderly people from the Honolulu Heart Program: a cohort study. Lancet 2001. 358:351-5

2 E Casiglia et al. Total cholesterol and mortality in the elderly. Journal of Internal Medicine 2003. 254:353-62

3 S Volpato et al. The inverse association between age and cholesterol levels among older patients: the role of poor health status. Gerontology 2001. 47:35-45

4 BW Schalk et al. Lower levels of serum albumin and total cholesterol and future decline in functional performance in older persons: the Longitudinal Aging Study Amsterdam. Age and Aging 2004. 33:266-72

5 DT Ko et al. Lipid-lowering therapy with statins in high-risk elderly patients.

JAMA 2004. 291:1864-70

Andrew Moore is honorary professor of health sciences at University College Swansea, and editor-in-chief of Bandolier

Henry McQuay is professor of pain relief at the Oxford pain relief unit and co-editor of Bandolier

Bandolier( is an independent monthly journal on evidence-based health care. Subscription costs £36 for 12 issues and subscribers receive the print journal three months before articles are available on the website. A subscription form is available on the website or from:

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say