“Women who drink two or more cups of coffee a day are less likely to get depressed,” said the BBC today, explaining that the caffeine in coffee may alter the brain’s chemistry.
The story comes from a study of over 50,000 women looking at whether those who drank more coffee were at less risk of getting depressed. It found that the more caffeinated coffee women drank, the lower their risk of developing depression. The same effect was not found for decaffeinated coffee.
This large study has some strengths but several limitations and is not robust evidence that coffee can prevent depression. It is possible the results are a case of ‘reverse causation’ and that the women who were depressed avoided drinking coffee. Also, it is possible that other factors such as family history or other circumstances influenced the risk of depression, although researchers tried to take account of these.
Overall, this study is not a reason to start drinking more coffee and further research is required to explore the possibility that caffeinated coffee may reduce the risk of depression.
Where did the story come from?
The study was carried out by researchers from Harvard School of Public Health, Brigham and Women’s Hospital, Harvard Medical School and Columbia University, US. It was funded by the National Institutes of Health.
The study was published in the peer-reviewed medical journal Archives of Internal Medicine.
Overall, the research was accurately reported by the papers and other media outlets. Both the BBC and The Telegraph pointed out the study has some limitations, the main one being that this type of observational study cannot prove cause and effect, in other words that coffee lowers risk of depression. The BBC also reported comments from an independent expert. The Mirror did not report any of the study’s limitations.
What kind of research was this?
This was a prospective cohort study that followed a total of 50,739 women for 10 years to find out if their intake of caffeine had any association with their risk of developing depression. This type of study is often used to investigate possible links between lifestyle interventions (such as coffee consumption) and health outcomes. The study was prospective and followed people over time. Therefore, is thought to be more reliable than a study in which researchers investigate lifestyle habits retrospectively or by questionnaire once the outcome (depression or not) is known.
The researchers point out that caffeine is the world’s most widely used stimulant and that 80% of caffeine is consumed through drinking coffee. They also say that previous studies in men have found that caffeine consumption decreases the risk of depression.
However, relatively few studies have examined this possible relationship. Furthermore, the possible association between caffeine use and the risk of depression, a chronic illness that affects twice as many women as men, is poorly understood.
What did the research involve?
Data from a large US cohort study was used to examine the possible association between caffeine and depression risk. The original research involved 121,700 American female nurses who were aged 30 to 55 when they enrolled in 1976. They provided researchers with updated information about their health and lifestyle every two years through mailed questionnaires.
The current study began in 1996 and looked at data on coffee consumption and depression from this date onwards. The researchers excluded women who could have had depression in the past, and anyone who had incomplete depression histories or whose data may have been incomplete or incorrect. This left them with 50,739 women, with an average age of 63 years, who were considered free of depressive symptoms at that time.
This group was followed up until 2006. Their consumption of coffee and other drinks, both caffeinated and non-caffeinated, was measured using validated questionnaires that the participants had completed every two years from 1980 through to 2004. Participants were asked about their coffee, tea, soft drink and chocolate consumption for the previous year.
The researchers classified participants into five categories of coffee drinking, ranging from one cup a week or less, to four cups a day or more. They used food composition data from official sources to calculate the amount of caffeine in a cup of coffee.
They then looked at whether the women had reported suffering depression from 1996 onwards. This was carried out through the questionnaire asking the women if they had been newly diagnosed with the condition by a doctor or had started using antidepressants regularly. This information was collected from 2000 and updated every two years until 2006.
The researchers also collected information about other factors that might have affected the results including lifestyle, medical history, age, weight, smoking status, exercise and social community group involvement.
Using the women’s reports of caffeine consumption, the researchers computed their average consumption of caffeine and other drinks. To investigate whether there was an association between caffeine consumption and depression, they allowed for a two-year ‘latency period’. For example, data on caffeine consumption from 1980 through to 1994 was used to look at new episodes of depression from 1996 to 1998, while data on consumption from 1980 to 1998 was used to look at new episodes from 2000 to 2002.
The analysis used standard statistical methods and the researchers adjusted their results for other factors that might affect the risk of depression, such as marital status, social involvement, smoking status, physical activity and other medical disorders.
What were the basic results?
During 10 years of follow-up (1996-2006), 2,607 new cases of depression were identified.
- women consuming two to three cups of caffeinated coffee daily had 15% less risk of depression (95% confidence interval [CI], 0.75 to 0.95), and those consuming four cups or more daily, had a 20% lower risk (95% CI 0.64 to 0.99) than women consuming one cup or less a day
- of the five caffeine consumption categories, women with the highest caffeine intake (500mg/d or more) had 20% less risk of depression than those consuming less than 100mg/day (95% CI, 0.68 to 0.95)
- decaffeinated coffee was not associated with depression risk
- there was no relationship between caffeine from non-coffee sources and depression risk
How did the researchers interpret the results?
The researchers say they found that depression risk decreased with increasing consumption of caffeinated coffee. They say that further investigations are needed to confirm this finding and to determine if caffeinated coffee can help to prevent depression.
The strengths of this well-conducted study include its large sample size, its prospective design and its use of a validated food frequency questionnaire, which was sent out seven times over 22 years.
However, the study had several limitations, as the authors acknowledge, which could affect its results. For example:
- It relied on women remembering and self-reporting their consumption of coffee and other drinks over the previous year.
- It also relied on women self-reporting their diagnoses of depression, rather than using other more reliable sources such as medical records.
- Although researchers tried to control for other factors (called confounders) that might influence the risk of depression, it is possible that some of these confounders were not taken into account and affected the results. It is possible that ‘reverse causation’ played a role in the results - in other words women who were depressed (but had not been diagnosed), might also be likely to drink less coffee. The authors tried to minimise this possibility by excluding at the start 10,280 women with severe depression. They also applied a two-year latency period when they computed the cumulative average of caffeinated and non-caffeinated drinks.
Overall, further research is required to explore the possibility that caffeine may reduce the risk of depression.
- Lucas M, Mirzaei F, Pan A, et al. Coffee, Caffeine, and Risk of Depression Among Women. Archives of Internal Medicine 2011; 171: 1571-1578