A US study showed that lonely people's cardiovascular systems worked differently to people who were not lonely, in ways which put them at higher risk of heart disease.

A second UK study suggests it may be wrong to link stress to heart disease.

They suggest previous research which made the link could have been weighted too heavily on people's own reports of their stress levels and symptoms, rather than objective measures.

Stressful tests

Researchers from the University of Chicago looked at why people who say they feel lonely have a higher heart disease risk.

They studied at 45 male and 44 female students from Ohio State University who filled in a questionnaire on loneliness and were given stress-inducing tasks including a mental arithmetic test, a writing test and having to give a speech to defend themselves against a false accusation of stealing.

The researchers, whose work is published in Psychomatic Medicine journal, measured blood pressure levels and vascular resistance - poor blood flow through vessels.

People who have a lot of different sources of contacts seem to do better as far as health is concerned

Andrew Steptoe, British Heart Foundation professor of cardiology

High blood pressure and vascular resistance have been linked to increased risk of heart disease.

In both groups, blood pressure rose before carrying out the tasks.

But the lonely students had higher vascular resistance.

In the other group, students' hearts beat faster - a normal reaction to stress.

In lonely students, this cardiac output was lower.

There was no difference between the lonely and non-lonely students in terms of habits such as drinking, smoking and diet, so researchers suggest the reason for increased risk for the lonely must lie elsewhere.

A second piece of research by the same team looked at six men and 19 women aged 53 to 78.

In those who said they were lonely, blood pressure was significantly higher in the older half of the group.

Amongst those who were not lonely, blood pressure was similar across all ages.

Different habits

In separate research, published in the BMJ, UK doctors say the suggestion psychological stress causes heart disease may have been "misleading".

The researchers looked at levels of stress reported by middle-aged Scottish men working in and around Glasgow in the early 1970s. who were followed for more than 20 years.

People who say they have more stress smoke more, drink more and take less exercise, but they still had better health

Dr John Macleod, University of Birmingham

Those who thought they were most stressed were also most likely to report symptoms of ill health, including angina which could even result in hospital admission.

They also drank more and took less exercise - but they also tended to have better jobs.

But objective measures of heart disease, such as electrocardiograms (ECGs) and death rates were actually lower amongst men reporting high stress.

They suggest it is unlikely that genuine coronary heart disease would not be associated with an increased death rates over 20 years.

The researchers, led by Dr John Macleod from the University of Birmingham say the crucial factor may be affluence.

He told BBC News Online: "Unhealthy behaviour is linked to high stress. People who say they have more stress smoke more, drink more and take less exercise, but they still had better health.

"But the bad effect of these things is offset by social advantage."

Professor Andrew Steptoe, British Heart Foundation professor of psychology at University College London, commenting on both pieces of research, said: "It is known that depression does affect health and on biological systems.

"The US study helps us think further about how social support is important for health and that people who have a lot of different sources of contacts - at work, with relations and friends - seem to do better as far as health is concerned."

But he said the did not think the BMJ paper was "as helpful as it could be".

"You really need to use more objective levels of stress.

"What most people think is important is the amount of control people have, which is not something which is measured here at all.

"People in high status groups to have demanding occupations and may feel quite stressed in one sense, but they're much more in control about how the work is done."