Medical practice

Tests to spot arthritis explored

“A simple blood test could help predict if someone is developing rheumatoid arthritis years before symptoms appear,” according to The Daily Telegraph. The newspaper says the test could enable patients to be treated earlier, helping to prevent some of the most devastating effects of the disease.

The news is based on a small study that examined blood samples from 86 patients with rheumatoid arthritis, taken before their symptoms developed. Researchers compared these samples with the blood make-up of 256 people who did not have the disease. They measured the levels of 30 substances that are linked to the immune system.

The possibility of identifying people with rheumatoid arthritis before they develop symptoms would be welcome as it may aid treatment to slow the disease. However, these 30 individual tests are probably not sensitive enough to do this yet. Further research must look at specific combinations of these tests.

Where did the story come from?

This study was conducted by Dr Heidi Kokkonen and colleagues from the Umeå University Hospital, and Department of Public Health and Clinical Medicine in Sweden. The study was supported by grants from several organisations, including the Swedish Research Council, the Swedish Rheumatism Association, and the European Community. The study was published in the peer-reviewed journal, Arthritis and Rheumatism.

The Daily Telegraph is one of the few papers to cover this story today. It provided a balanced report of the main findings from the study, and relevant quotes from experts.

What kind of research was this?

In this exploratory study, the researchers tested blood samples to see if they could find indicators of the future development of rheumatoid arthritis (RA). These blood samples were taken from individuals before and after they developed symptoms of RA, and from control subjects without the condition.

The researchers analysed the blood samples for a range of chemical messengers: cytokines, cytokine-related factors and chemokines. These substances are secreted by the immune system, and act to carry signals locally between cells.

Although small, this case-control study provides some hope for people living with RA. However, it will need to be repeated in order to examine in more detail which precise bloods test were most useful and how accurate they are. There is also a need for further studies looking at how asymptomatic people likely to develop RA could be targeted for these types of tests.

What did the research involve?

The researchers designed a nested case-control study. In this type of study, researchers collect a sample of patients with and without a disease from a larger population-based cohort. In this case, they drew participants from the Biobank study, which had taken blood tests since 1985. The participants from the Biobank study were all taken from the adult population of the county of Västerbotten in northern Sweden, who have been continuously invited to participate in the study. There have been several publications from this Biobank study already.

For their case-control study, the researchers selected patients with RA who fulfilled the American College of Rheumatology Classification criteria for RA, and who also knew the onset date of their symptoms of joint disease (the cases). From these, they found 86 people (65 females and 21 males) who had donated blood samples before the onset of any symptoms of joint disease. They matched them with 256 people who had also donated blood but did not have the disease (the controls). On average, the cases in the sample had given blood about 3.3 years before any symptoms of the disease.

The researchers measured the levels of 30 cytokines, related factors and chemokines in the blood samples, using standard lab-based techniques. However, these testing techniques are not yet in clinical practice. They then used advanced modelling techniques (Random Forest modelling) to analyse the associations between all these substances and the presence of RA. This analysis adjusted for the influence of smoking, sex and some genotypes.

Finally, the researchers translated their results into sensitivities and specificities, two measures of the diagnostic accuracy of the test.

What were the basic results?

Compared with the controls, several of the substances tested were significantly raised in the cases before the onset of their RA. These raised substances were linked to signs of general immune activation, specific immune activation and regulatory pathways. The levels of these substances were especially raised in individuals who were positive for rheumatoid factor.

The single test with the highest sensitivity for predicting RA was for one measuring a substance called ‘eotaxin’. When used alone, the test had a sensitivity of 22.4% at a pre-set specificity of 95.3%. This means that only about one in four people who go on to develop RA will test positive with this test. Sensitivity improved when all 30 tests were used together.

How did the researchers interpret the results?

The researchers conclude that people who went on to develop RA had significantly increased levels of several cytokines, related factors and chemokines prior to their symptoms. They say that the types of substances elevated before disease development suggests that at this stage the immune system is ‘adaptive’ (i.e. still responding to the triggers of rheumatoid arthritis). After the disease has started, the involvement of the immune system is more general and widespread.

They say that their findings present an opportunity for “better prediction of the risk of developing RA and, therefore, possibly preventing the disease progress”.


This was a small exploratory study that will be of interest to researchers in the field. The authors acknowledge some limitations, such as the small sample size: only a few patients had given samples both before and after the onset on RA. This means that there was only a small body of data to examine for the associations.

The tests themselves were not very sensitive when used on their own, and provided a lot of false negative results (i.e. failing to detect the presence of the condition). It remains to be seen whether it will be better to test for all these substances (which seemed to increase sensitivity) or whether there is a combination of fewer tests that are highly sensitive and specific.

The researchers also speculate about the biological processes in people with RA prior to symptoms, suggesting what might cause some of the elevation in cytokines and other substances. But these theories will need to be tested in further studies.

It is too soon to say whether these sorts of tests can become standard practice, or for whom they will be most useful. However, a combination of tests seems more promising for predicting people who will develop RA. This type of study does provide some direction for future research.

NHS Attribution