Looking forward to ill health? by Professor Peter Thomas

The psychological impact of having a chronic disease such as MS or diabetes is difficult to ignore. Individuals respond differently, but often they will have to readjust their life goals and expectations (possibly against a background of uncertainty about the course of their disease), learn to adjust to changes in mobility and function, have difficulties in maintaining social networks and close family relationships, and have to cope with long term symptoms such as pain, stiffness, and fatigue. Some people will have difficulty doing this, and might develop psychiatric symptoms such as depression or anxiety. In others the psychological effects may be more subtle although still significant. Further, any impact could start to snowball if disease progression or symptoms become worse in people for whom the psychological impact is high. Many people with chronic disease are helped by medical treatments, but when a cure is not an option and when good disease/symptom management cannot be achieved, psychosocial issues become increasingly important. In some diseases, psychological interventions such as cognitive behavioural therapy have been shown to be helpful.  

Another intriguing angle concerns the role that psychological factors might play in the aetiology of chronic disease. The role of stress and personality in heart disease and, to a lesser extent, depression in cancer has been studied before. Such links are biologically plausible via immune and endocrine pathways, and experimental work has supported this. The role (if any) of these, and other, psychological factors in many other chronic diseases is less clear.

A methodological feature of much of the research that looks at the link between psychological factors and chronic disease is the use of cross-sectional studies. Such studies capture a snap shot in time. They are extremely useful for investigating associations between psychological factors and chronic disease, and have the advantage that they are relatively simple to do. The major disadvantage is that they do not allow one to disentangle whether any differences in psychological factor are a result of having chronic disease or whether they precede the chronic disease and possibly involved in aetiology. Looking at a photograph of a boy standing beside a football does not help one understand what happened prior to the photograph or following the photograph. Is the ball in the picture just by chance? Did the boy just place the ball at his feet? Has someone just passed the ball to the boy? Is the boy about to pass the ball to somebody else? Similarly a cross-sectional study does not allow one to distinguish whether the psychological differences occurred prior to or subsequent to the onset of chronic disease.  

Continuing the image-capturing analogy, video footage is much better at helping one understand a sequence of events because it includes a time dimension. The video footage would show how the ball got there and what happened to it next. In research this is done by using prospective, longitudinal studies. To ascertain whether psychological factors preceded the onset of chronic disease one could identify a disease-free group of individuals, make the psychological measurements and then follow them up a number of years later to see who developed chronic disease. To ascertain whether psychological factors occurred after the development of chronic disease, one might consider a group of individuals newly diagnosed with the disease and a group of people without the disease and then follow them up a number of years later to make the psychological assessments.  

Longitudinal designs take time into account and allow one to distinguish "causes" from "effects". Although there are considerable advantages to doing longitudinal research, there are obvious practical difficulties that have to be faced. Taking the example of following up disease free individuals to see who develops chronic disease, follow-up might have to be done over a long period of time, maybe 10 or 20 years. If the chronic disease is rare, then large numbers of individuals will need to be studied in order to identify a reasonable number with chronic disease. The RDSU is currently examining these research questions using secondary data analysis of existing large prospective longitudinal studies that originally collected the information for completely different purposes.  

Researchers have tried to bypass these problems. For example, following up individuals at high risk of chronic disease, perhaps because of family history, might mean that sample size can be smaller and follow-up shorter. "Quasi prospective" designs have also been used. An example is measuring psychological variables in women being routinely screened for breast cancer, and in whom those with and those without breast cancer will be known within weeks. Although a neat solution, this design has been criticised. Some women may well suspect that they have breast cancer before the scan, for example breast changes or lumps that they have not yet acted upon. In these women the time sequence of psychological changes and development of chronic disease will not be clear cut.  

The research that the RDSU is doing on emotional processing has been described in a previous issue of RADAR. Measuring emotional processing in healthy people and people with chronic disease has revealed fascinating differences in the way the groups  process their emotions, and these differences vary according to the type of chronic disease. This work has been done using cross-sectional research designs, and so the interpretation must be guarded. Further work is planned to collect data longitudinally to determine whether deficits in emotional processing precede the development of chronic disease, or are a consequence of having a chronic disease. 

Peter Thomas
Professor of Health Care Statistics and Epidemiology
RDSU