After stress comes cancer

An analysis of more than 160,000 health survey participants from England and Scotland published in the British Medical Journal finds a link between cancer mortality and stress. Even when classical risk factors are included in the calculation, the risk of dying from cancer increases fourfold in some cases in psychologically stressed individuals. A clear signal that stress not only has negative cardiovascular effects.

(ag) Data from a total of 163 363 women and men who were at least 16 years old (mean 46 years) were analyzed or pooled. This was based on 16 prospective cohort studies from 1994 to 2008 (mostly from England). Individuals had completed annual stress scores based on the 12 questions of the General Health Questionnaire (GHQ-12) and consented to have their medical records used for research purposes. The GHQ-12 is a widely used tool for such measurement and consists of various questions such as:

Have you slept less in recent weeks because of worries?
Or: Have you felt a lack of confidence in recent weeks?

Here, “no, not at all” and “no more than usual” resulted in a score of 0 and “more than usual” and “much more than usual” resulted in a score of 1.

The analysis of the English data was not about the incidence of cancer, but about mortality. Consequently, we looked at whether psychological stress (anxious and depressive symptoms), independent of other factors, increased the chance of dying from cancer – rather than other deaths. Only the three Scottish studies had also surveyed the first occurrence or diagnosis of such disease, i.e., incidences.

Clear risk association

A total of 16,267 deaths occurred, slightly more than a quarter of them due to cancer. The mean observation period was approximately nine years. They now included other factors such as age, education, BMI, alcohol and tobacco consumption in the multivariate analysis and came to the following conclusion: compared to individuals who were below 7 on the GHQ-12 stress level, those with 7 or more possible units (this group the researchers defined as “highly symptomatic”) had one:

26% increased risk of dying from any cancer (95% CI 1.11-1.42).
84% increased risk of dying from colorectal cancer
increased risk by a factor of 2.42,
to die from prostate cancer
increased risk by a factor of 2.76,
to die from pancreatic cancer
risk increased by a factor of 2.59,
to die from esophageal cancer
increased risk by a factor of 3.86,
to die of leukemia.

For prostate and colorectal cancer, there was even a “dose-response relationship”: the risk increased gradually with each increase on the stress scale.

Have we forgotten anything?

Other potentially confounding factors for cancer mortality, such as deprivation (no access to high-quality medical treatment), were not included in all of the studies mentioned, which is why the researchers could only perform a subgroup analysis here. However, this hardly changed the result of the main analysis, i.e. the association between stress and cancer mortality.

Furthermore, it was examined whether a reverse causal direction might not be present: Were people with undiagnosed cancer included in the study who already felt certain effects of the cancer, such as fatigue or even pain, and thus experienced stress or mistook such effects for symptoms of stress? This possibility was ruled out by excluding people who died of cancer as early as five years after study inclusion in a further subgroup analysis, but this also had little effect on the result.

Incidence increased

The three Scottish studies showed that the incidence of cancer was also increased overall by 16% for the “highly symptomatic” group with a stress score of 7-12. The authors note, however, that the associations are weaker here, which is due not least to the small number of cases.

Stress not only has a cardiovascular effect

In recent years, the link between mental and physical health has attracted attention, especially with regard to cardiovascular risk factors. Stress not only increases blood pressure, but has negative effects on our cardiovascular system far beyond that: psychosocial risk factors such as acute emotional or chronic social stress, negative affect, certain personality factors, and fatigue states lead to comparable increases in risk for coronary heart disease (CHD) as established parameters (these include smoking, diabetes, obesity, passivity). According to a meta-analysis [1], outbursts of anger cause an almost fivefold increase in the risk of myocardial infarction/ACS. The critical time interval is two hours after such an outburst of anger. Apparently, there is also a dose-effect relationship here: the more trouble, the higher the danger.

The association between stress and cancer, on the other hand, has been the subject of few studies. The results of this study suggest that stress can have an unfavorable effect on cancer or even cause it to develop in the first place via multiple immunological, inflammatory, and hormonal mechanisms. It is known that stressed individuals also fulfill other risk profiles more frequently: There are more smokers and overweight people among them, and they maintain an overall unhealthier lifestyle (irregular and unhealthy diet, little exercise, alcohol consumption). However, some of these factors could be controlled in the study without changing the outcome to any relevant extent.

What now?

The associations and causal directions between stress, lifestyle and cancer risk remain highly complex – the study therefore primarily shows that they are worth exploring in more detail. Concrete prevention efforts cannot (yet) be derived from this: Should stress in society in general also be addressed in terms of cancer prevention, or is it enough to avoid the classic factors such as smoking, little exercise and an unhealthy diet in order to prevent or reduce both cancer and stress?

Source: Batty GD, et al: Psychological distress in relation to site-specific cancer mortality: pooling of unpublished data from 16 prospective cohort studies. BMJ 2017; 356: j108.

Literature: