Unpleasant or demanding interpersonal encounters increase hypertension risk among older adults, a new study has claimed.

The study by Carnegie Mellon University's Rodlescia Sneed and Sheldon Cohen provides some of the first concrete evidence that negative social interactions not only influence psychological well-being but also physical health - in this case, blood pressure levels.

Hypertension affects an estimated 65 million Americans and is a major contributor to cardiovascular disease, the leading cause of death in the US.

For the study, Sneed and Cohen used data from the Health and Retirement Study, a multi-year survey of 1,502 healthy adults aged 50 and over.

In 2006, the frequency of negative interactions - exchanges or behaviours that involved excessive demands, criticism, disappointment or other unpleasantness - with their partners, children, other family members and friends was assessed by questionnaire.

Blood pressure was measured at this assessment as well as four years later.

The results showed that each increase in the total average negative social interaction score was associated with a 38 per cent increased chance of developing hypertension over the four-year period.

Younger older adults - those aged 51-64 - were also more affected than those 65 or older.

The researchers also observed sex differences in their findings. While negative interactions predicted hypertension risk among women, these interactions were not related to hypertension risk among men.

"There is a body of evidence in social psychology research suggesting that women care more about and pay more attention to the quality of their relationships," said Sneed, a Ph.D candidate in psychology.

"Our findings suggest that women are particularly sensitive to negative interactions, which is consistent with this previous work," Sneed said.

The researchers also found that the type of relationship matters. Negative interactions between friends and family led to an increase in hypertension risk while poor encounters with partners and children did not make a difference.

The study was published in the American Psychological Association's journal Health Psychology.