Richard Rahe, J L Mahan & R J Arthur 1970

AIMS: During the 1950s many studies indicated a relationship between stress and poor health. One of the most important of these was the 1958 study by Ray Rosenman & Meyer Friedman which found a link between the kind of stress experienced by ‘Type A’ personalities and coronary heart disease  - see Stress & Vulnerability.

Rahe, Mahan & Arthur were more interested in environmental factors which caused stress to the point of illness, rather than temperament.

The idea of ‘life events’ causing stress to the point of illness had begun in 1919 with the ‘life chart’ work of Adolph Meyer. His work became the foundation for the Schedule of Recent Events (SRE) developed by David Hawkins, R Davies & Thomas Holmes (1957); this looked at the cumulative effect of life events in causing stress. In 1967 Holmes & Rahe added the idea of the magnitude of different life events - measured in ‘life change units’ (LCUs) - to get a more precise understanding of the cumulative effect. They examined the medical records of over 5,000 medical patients as a way to determine whether stressful events might cause illnesses. Patients were asked to tally a list of 43 life events based on a relative score. A positive correlation of 0.3 was found between their life events and their illnesses. And thus, the SRE became the Social Readjustment Rating Scale (SRRS).

Most studies into the relationship between stress and illness in the 1950s and 1960s - eg: William Greene (1954), Greene, Lawrence Young & Scott Swisher (1956), Rahe, M Meyer, M Smith, G Kjaer & Holmes - were carried out retrospectively - ie: the participants were already hospitalised with illness. So the participants’ memory of stress factors which might have contributed to their illness may have been coloured by the very fact that they were now ill.

Rahe, Mahan & Arthur wanted to conduct a prospective study - ie: to predict illness as a result of stress factors.

Thus, the researchers’ aim was to test whether participants who scored highly on the SRRS would become ill - in which illness through stress could be deemed predictable.

PROCEDURE (METHOD): 2,684 American naval men of varying ranks were asked to complete an SRRS specially customised for navy personnel The questionnaire was used to assess the stressful life events each man had experienced over 4 consecutive 6-month periods in the lead-up to their deployment on one of 3 navy cruisers. Each stressful factor was assigned an LCU weighting that indicated the severity of the source of stress. A total LCU score was calculated for each of the 4 6-month periods prior to embarkation.

The sample represented 90-97% of each ship’s crew and fell to a total of (less than 10%) due to transfer off the ships. The average age of the men was 22.3 years; approximately 2/3 of them were high school graduates; and they ranged in experience from apprentice seamen to high-ranking officers with 30 years’ naval experience.

The ships were on 6 months active duty and 2 of them were on military missions off the coast of Vietnam.

On board any illness, no matter how trivial, had to be recorded by the ship’s medical facility. When the data on health was analysed, any sickness believed to be motivated by a desire to avoid work was discarded - as were any reports of pre-existing conditions. The remaining information on the number, type and severity of new health problems was used to produce an illness criterion.

Neither the participants themselves nor the ship’s medical facilities knew their health records were going to be used in this way.

FINDINGS (RESULTS): The only significant correlation found - r = 0.118 - was between the LCU total for the 6 months prior to departure and the illness score. Consequently only the data relating to the final pre-departure 6 months was used in subsequent analyses.

Overall totals were calculated by putting participants in order of highest to lowest LCU and dividing the list into 10 bands (each 1/10 of the ship’s crew). When divided up in this way, 2 of the ships showed large differences but one was not so great. So the data for all 3 vessels were combined which obscured the variations between the third ship and the others. The only big differences then were at each end of

The standard Social Readjustment Rating Scale

the scale. To overcome this, the grouping was changed, combining some of the 10 bands to make 4 bigger groups.

The final stage of analysis was to re-group the men’s LCU scores into bands 0-99, 100-199, 200-299, 300-399, 400-499, 500-599, 600-600+ and plot these against mean illness rates. This produced a linear relationship between exposure to life events and ill health.

The 2 ships judged to have the easiest missions showed the strongest relationships between LCU totals and mean illness rates. It was considered possible that the current stress levels on the third ship may have nullified the effects of pre-embarkation stress factors.

CONCLUSIONS: Rahe et al concluded that there was a link between the higher LCUs from the 6 months prior to departure and higher illness rates on board ship. When the  pre-departure life changes were low, so where the onboard illness rates. However, the study indicated this relationship could be masked by onboard stressful experiences which increased the illness rate for the entire crew.

Additionally, the link was stronger for older men (over 21s) and married men.

CRITICISMS (EVALUATION): The study had a number of strengths:-

• It was a prospective study - ie: the reporting of stress factors was not influenced as a result of already being ill as a consequence of those stress factors.

• Neither the participants nor the ship’s medical staff knew the records were going to be used in the study - therefore, the reporting of illness and the recording of it can be considered to have been unaffected by any form of demand characteristics.
• The crew of each ship would have been exposed to the same current environmental factors - eg: similar pathogens and food - and to have received similar health care.
• It was customary for crew members to report even minor illnesses and the medical facilities to report them according to procedure - thus,ensuring consistent, highly-detailed records of the participants’ state of health
• It was a large-scale study, representing a wide demographic spread.
• The reliability and validity of the SSRS has been improved by several revisions. J Mendels & N Weinstein (1972) evaluated the LCUs allocated to each event and found a high correlation between their findings and those of Holmes & Rahe (1967) even though they used a sample which was different in both age and education.

• Another prospective study by M A Gupta & Aditya Gupta (2004) supported Rahe et al in terms of both reliability and validity - the Guptas finding a significant correlation between life events causing stress and the deterioration of skin disorders (such as scalp itching).

However, the study also has a number of weaknesses:-

• The correlational coefficients found by Rahe et al - eg: r = 0.118 - were all quite low, only the sheer size of the sample group making such low correlations statistically significant. In other words, Rahe et al could only find a weak association between life events and stress to the point of illness. Cause-and-effect most definitely cannot be inferred from this study.
• The study is open to accusations of cultural bias - the participants were all American - and gender bias - the participants were all men. The sample may not even have been representative of American sailors as it was not a random sample.
• The SRRS is a self-report measure which leaves it open to all kinds of cognitive biases. For example, men about to go to war may feel they should answer in a certain way (social desirability bias).  A 1991 study by Karen Raphael, Marylene Cloitre & Bruce Dohrenwend illustrates the inherent dangers of self-reporting: when women, asked to report stressful events each month over a 10-month period, were asked to recall the events again at the conclusion of the study, only 1/4 of the originally-reported categories appeared.
• The study - indeed, the concept of the SRRS - does not allow for individual differences. For example, some of the sailors might have looked upon a life event - say, impending divorce, as an opportunity rather than a threat.
• The study has some serious ethical issues - with the privacy and confidentiality (health records) of the navy personnel breached, no right to withdraw granted and no opportunity for informed consent - thought they did give consent to completing the SRRS questionnaires.