A hearty cup of tea?

Monday June 21 2010

“People who drink several cups of tea or coffee a day could be at lower risk of heart disease,” reports the BBC. The news channel says drinking tea could cut the chances of a heart attack by up to a third.

The news is based on a Dutch study that followed 38,000 people for over a decade, looking at their risk of stroke, cardiovascular incidents and death. When examining the association between coffee consumption and events relating to heart disease, the study found that drinking moderate amounts of coffee (two to three cups a day) was better than a little or a lot. With tea, drinking more than six cups a day was linked to the lowest risk.

There have been a number of contradictory research findings about the benefits and harms of caffeine. (For example, the Metro reports that a separate Italian study has recently found an increased risk of rheumatoid arthritis for women who drink tea.) Although the Dutch study is an interesting contribution to the debate, its limitations mean that it cannot prove that coffee and tea have a direct effect on heart attack risk.

Where did the story come from?

The study was carried out by researchers from the University Medical Center Utrecht and the National Institute for Public Health and Environment, both in the Netherlands. It was funded by the European Commission, the Dutch Cancer Society, the World Cancer Research Fund, the Netherlands Organisation for Health Research and Development and the Dutch Ministry of Public Health, Welfare and Sports.

The study was published in the peer-reviewed Journal of the American Heart Association.

The newspapers have covered this story well, with many explaining the results in the wider context of better established evidence about reducing coronary heart disease risk. A senior cardiac nurse at the British Heart Foundation is quoted as follows by several sources: “It's worth remembering that leading a healthy overall lifestyle is the thing that really matters when it comes to keeping your heart in top condition. Having a cigarette with your coffee could completely cancel any benefits, while drinking lots of tea in front of the TV for hours on end without exercising is unlikely to offer your heart much protection at all."

What kind of research was this?

This large cohort study assessed the effects of tea and coffee consumption on health outcomes over a 13-year period, specifically any effects on cardiovascular health. The researchers have attempted to adjust their calculations to account for some potentially confounding factors (that may link the consumption of tea and coffee with adverse outcomes). As with all observational studies, there is a concern over whether all possible confounding factors have been addressed or fully adjusted for. The researchers have discussed some of the potential shortcomings of their study.

What did the research involve?

The researchers highlight the controversy that surrounds the benefits and harms of coffee consumption and note that the association between coffee and cardiovascular diseases remains controversial. They also say that for tea consumption, a systematic review assessing the association with coronary heart disease and stroke are inconclusive, but in the European region, tea consumption appears beneficial in relation to myocardial infarction.

In this study, researchers investigated the relationship between tea and coffee consumption and cardiovascular disease in a large cohort of healthy Dutch men and women. There were 37,514 participants, who were also participating in one of two other cohort studies and who were recruited from 1993 to 1997. The first study enrolled women aged 50 to 69 who were participating in a breast screening programme and the second was in men and women aged 20 to 65. The researchers excluded all those who had missing information about tea and coffee consumption or had cardiovascular disease at baseline (the start of the study).

At baseline, the participants completed a questionnaire asking about chronic diseases, presence of a variety of potential risk factors, demographics and their medical and lifestyle histories. Their height, weight, hip-and-waist circumference and blood pressure were measured. Physical activity was assessed according to a validated physical activity index.

Participants were also given a food frequency questionnaire that assessed their average daily consumption of 178 different foods during the previous year, including how many cups of coffee or tea they regularly drank during the past year and which types of coffee (regular, decaffeinated, other etc). Tea and coffee consumption was then divided into six ranges (less than one cup per day, one to two, two to three, three to four, four to six and more than six cups per day). Some of these categories were collapsed during analyses due to the small number of people within each group.

The researchers then noted the health outcomes of the participants up to 13 years after the baseline questionnaires, specifically any events or deaths from coronary heart disease (CHD) and stroke. They also analysed the combined outcome of morbidity and mortality (events plus deaths) for stroke, CHD and death due to any cause. Their study then assessed whether there was an association between the different levels of consumption of coffee and tea and the negative health outcomes, while taking into account a number of possible confounding factors. The confounders adjusted for were age, gender, education, physical activity, smoking, waist circumference, menopausal status and HRT use, alcohol intake, total energy intake and intake of saturated fat, fibre, vitamin C and total fluid intake.

What were the basic results?

Over the course of follow-up, 1,950 cardiovascular events occurred (563 from stroke, and 1,387 from coronary heart disease (CHD)). There were 1,405 deaths (including 70 from stroke and 123 from coronary heart disease).

The researchers found the lowest risk of CHD events to be linked with drinking more than two but fewer than three cups of caffeinated coffee per day. The increased risk of stroke with more than six cups of coffee per day was no longer significant once the researchers had adjusted for confounding factors. After taking into account these confounders, coffee consumption was not associated with death from stroke, death from any cause or death from coronary heart disease (although the researchers state that “although not significant, coffee slightly reduced the risk for CHD mortality”).

For tea, consumption of more than six cups per day was associated with the lowest risk of CHD events (HR 0.64, 95% CI 0.46 to 0.90, p=0.02). The relationship between tea and stroke risk was not significant after adjusting for confounders. There was a significant link between tea consumption and death from CHD, with the lowest risk of CHD death linked to two of the consumption ranges: one to three cups per day and more than three but fewer than six cups per day. After adjusting for confounding factors, there was no significant link between tea consumption and deaths from stroke or due to any cause.

How did the researchers interpret the results?

The researchers say that “high tea consumption is associated with a reduced risk of CHD mortality”. They note that their results suggest “a slight risk reduction for CHD mortality with moderate coffee consumption” and “strengthen the evidence on the lower risk of CHD with coffee and tea consumption”.

For coffee, the relationship with CHD events was “U-shaped”, i.e. higher risk with both very low and very high consumption. For tea, there was a linear inverse association (reducing risk with increasing consumption).


This large cohort study with a long follow-up time has found an association between some levels of tea and coffee consumption and reduced risk of coronary heart disease events. The study authors highlight the following important limitations of their research, many of which are relevant because of the study design:

  • A relatively small number of people died during the course of follow-up (123 from CHD and 70 from stroke). When considering these particular outcomes, the small numbers seen do not provide much statistical power to detect differences between the different consumption groups.
  • The researchers relied on participants to recall their tea and coffee consumption over the course of a year at baseline. There are two potential problems with this. Recall may not be 100% accurate and collecting information on consumption only at baseline does not take into account very likely changes in consumption patterns over time.
  • Assumptions were made about the type of tea consumed (as this was not specified in the baseline questionnaires); such as the majority of tea consumed was black tea.
  • Importantly, they note that they cannot exclude the possibility that some factors were commonly linked with the exposure (tea and coffee consumption) and with the outcome (CHD events). In particular they say that coffee drinkers tend to smoke more and have less healthy lifestyles than tea drinkers and that this could explain higher risk of adverse outcomes in people consuming a lot of coffee. While they have adjusted for some lifestyle factors, they acknowledge that they may not have done this fully.
  • While they acknowledge that adjustment for the presence of diabetes, high blood pressure and high cholesterol did not alter the associations, this may be because of the crude way they measured the presence of these diseases (through self-reporting at baseline).

Overall, the limitations and design of this study means that it adds more to the discussion about the benefits and harms of caffeine, but cannot be taken as proof that tea or coffee cause reductions in the risk of heart disease. There are well-established ways of reducing the risk of heart disease, including healthy eating and physical activity, rather than relying on high tea consumption or moderate coffee intake. The food standards agency makes particular recommendations about caffeine consumption for pregnant women (recommending no more than 200mg of caffeine per day, which is approximately two mugs of instant coffee or tea).

Analysis by Bazian
Edited by NHS Choices