Mental health

Stress and heart attacks

“Stress increases the chances of dying from heart-related problems by five-fold,” according to The Daily Telegraph. The newspaper reported that a study of people aged over 65 found that those with high levels of stress hormones were five times more likely to die from cardiovascular causes, such as heart attacks and strokes, within the next six years.

The study measured levels of a stress hormone, called cortisol, in the urine of 861 older adults and tracked their records to identify any deaths and their causes. Those with the highest cortisol levels had an increased risk of death from cardiovascular disease, but not from other conditions. However, only 41 people from the whole study died from cardiovascular disease, which limits the reliability of the study’s findings. Also, the influence of important factors, such as diet and physical activity, was not fully accounted for and may have affected the results.

Overall, these limitations mean that the results should be interpreted cautiously and will need confirmation in larger studies.

Where did the story come from?

The study was carried out by researchers from VU University Medical Center in the Netherlands and other research centres in the US and Italy. It was funded by the Italian Ministry of Health, the US National Institute on Aging and the US National Heart, Lung and Blood Institute. The study was published in the peer-reviewed Journal of Clinical Endocrinology and Metabolism.

The Daily Telegraph and BBC News both provided balanced coverage of this study.

What kind of research was this?

This prospective cohort study looked at whether there was a link between levels of the stress hormone cortisol in urine and risk of death in older people.

The body releases the hormone cortisol when it is under physical or psychological stress. It has various effects on the body, including increasing blood pressure and the amount of blood pumped out by the heart. Cortisol is meant to help the body recover from stress, but if it remains at high levels over a long time, it may cause damage to the body. However, the researchers say that there is little direct evidence that raised cortisol levels can increase the risk of cardiovascular events, such as heart attacks.

This type of study, which assessed people’s cortisol levels at the start of the study and followed them up over time to see who died, is the best way to look for a link between cortisol levels and risk of death.

What did the research involve?

This study used data from the InCHIANTI study, which looked at various aspects of health in elderly people. From 1998 to 2000, researchers enrolled people aged 65 years and over (average age 75) to take part. Participants gave urine samples, consisting of all the urine passed in a 24-hour period, which were tested for levels of the stress hormone cortisol. The participants were then followed up for six years to see who died in this period.

The researchers then looked at whether the level of cortisol at the start of the study predicted how likely an individual was to die during the follow-up period due to any cause or from cardiovascular disease.

The urine samples were collected for an entire day (24 hours). People taking drugs or with medical conditions which could affect cortisol levels were excluded, as were people who did not provide a full 24-hour urine sample. This left 861 people who were eligible for inclusion in the current analyses. Deaths were identified using the mortality registry of the region where the study was being performed (Tuscany in Italy). The researchers then obtained the death certificates of those who died to identify their cause of death.

Deaths were separated into deaths from cardiovascular causes (such as heart attacks or strokes) and all other deaths. The researchers then compared the risk of death from these causes in people with the highest levels of cortisol (top third of the measurements, more than 111 micrograms) and the lowest levels of cortisol (bottom third of the measurements, less than 78 micrograms).

The researchers also looked at the effect of each 48-microgram increase in cortisol levels (48 micrograms being the standard deviation of the range of cortisol measurements).

At the start of the study, participants had provided information about themselves, including their medical history and drug treatments. They also had a medical examination and filled in questionnaires assessing depressive symptoms and cognitive ability. Hospital discharge records were also obtained for the participants to help identify any existing cardiovascular disease.

In their analysis, the researchers took into account factors that could affect their results, including the participants having cardiovascular disease at the start of the study, how many other chronic diseases they had, their age, gender, years of education, smoking, current alcohol intake, body mass index, waist circumference, blood pressure, depressive symptoms and cognitive ability.

What were the basic results?

During follow-up, 183 out of the 861 participants died (21.3%). Of the total number of participants, 4.7% (41 people) died from cardiovascular causes.

There was a link between the level of cortisol and risk of death from any cause. Participants with the highest cortisol levels had a 74% greater risk of death during follow-up than those with the lowest levels (hazard ratio [HR] 1.74, 95% confidence interval [CI] 1.15 to 3.62).

The level of cortisol in the urine at the start of the study predicted the likelihood of death from cardiovascular causes but not from non-cardiovascular causes. People in the group with the highest levels of cortisol in their urine were five times more likely to die from cardiovascular disease during follow-up than those in the group with the lowest levels (HR 5.00, 95% CI 2.02 to 12.37).

For every 48-microgram increase in cortisol levels, there was a 42% increase in risk of death from cardiovascular disease during follow-up (HR 1.42, 95% CI 1.06 to 1.90). This relationship appeared to be the same in people who had cardiovascular disease at the start of the study and those who did not.

How did the researchers interpret the results?

The researchers concluded that: “high cortisol levels strongly predict cardiovascular death among persons both with and without pre-existing cardiovascular disease.”


This study suggests that urinary cortisol levels in older adults may be related to risk of death from cardiovascular causes. However, there are some limitations to consider:

  • As with all studies of this type, it is not possible to say for certain whether higher cortisol levels cause the increased risk of cardiovascular death, or whether other factors have an effect. While the researchers’ calculations were adjusted for a number of these factors, their effects may not have been wholly removed, and other unmeasured factors, such as diet and physical activity level, may have had an effect.
  • There were few deaths due to cardiovascular disease (41 deaths) in this study, which lessens the reliability of the findings. Analyses of this outcome should be interpreted cautiously.
  • The study measured urinary cortisol over 24 hours, which is probably better than measuring a single sample. However, even a 24-hour sample may not be representative of a person’s average cortisol levels in the long term.
  • The study included some people who already had cardiovascular disease (13% of the participants). In these people, it would not be possible to say whether cortisol levels played a role in their prior development of these conditions, as they were only measured after the onset of the disease. Although statistical tests suggested that the presence of cardiovascular disease at the start of the study did not affect the results, it may have been better to exclude people who already had cardiovascular disease at the start of the study.
  • Although cortisol is considered to be a marker of stress, this study did not assess how stressed individuals were feeling. Therefore, it cannot tell us about the link between feeling stressed and the risk of death.

Overall, further larger studies are needed to confirm these findings.

NHS Attribution