The Royal College of General Practitioner's (RCGP's) Oral Contraception Study was set up to explore the longterm health effects of oral contraception.1 The recruitment of 47 000 women by 1400 GPs in the late 1960s was a monumental achievement. Not only has the Oral Contraception Study achieved its original objectives but, given the size of the cohort, it has proven to be a valuable resource to study many diverse issues, including the prevalence of chronic pain and the relationship of tubal sterilisation and subsequent allcause death.
By 1994 a quarter of the women from the original cohort were still under observation and were invited to complete a questionnaire which included questions on their current health status and lifestyle. With a response rate of 85% this cross-sectional survey, combined with baseline data from the cohort study, has enabled Iversen and colleagues2 to investigate the relationship between combinations of risk factors and all cause mortality in women. The study, published in this month's edition of the BJGP, focuses on four modifiable lifestyle risk factors: smoking, alcohol consumption (excess alcohol intake or never drinking), physical inactivity, and body mass index outside the normal range (BMI<18.50 or BMI >25.00). The researchers demonstrate that women with multiple lifestyle risk factors had higher mortality risks than those reporting none. Assuming causality and reversibility, they estimate that 60% of deaths of women in this cohort may have been prevented by the avoidance of all four of the modifiable lifestyle risk factors considered.
Examining the study's strengths and limitations, this study is a useful opportunistic exploitation of existing data sets. The weaknesses of utilising this cohort for further studies other than the primary purpose and the conduct of a cross-sectional study within a cohort study have been rehearsed on previous occasions. Similarly the lack of generalisability of the RCGP Oral Contraception Study cohort, which consists largely of white European women who in 1968 were married or living as married, is widely acknowledged.
Relating to this particular study there are issues about the summary classification of low and elevated BMI measurements and the lack of detail of social class (manual or non-manual). The analysis is confined to only four lifestyle risk factors and is based on a single snapshot of lifestyle in middle life, as the researchers were unable to ascertain a full life history of the risk factors. Nonetheless, Iversen et al's recent observations are consistent with those of other UK and US cohort studies of lifestyle factor combinations on mortality. This paper serves as a timely reminder of the relationships between lifestyle and mortality, and re-emphasises the importance of reducing smoking and promoting activity.
While the limitations of the data and methodology of this study will no doubt stimulate debate, more challenging questions emerge as one considers the role of primary care in addressing these epidemiological observations. Iversen et al's study defines and quantifies a problem, but it does not give us solutions to lifestyle modification. Although lifestyle is influenced by a huge range of personal, societal, and environmental factors, general practice has a potentially important role to play, given the 75% population coverage it provides within a single year. However, as primary care clinicians we still have relatively few robust and effective lifestyle interventions in our armamentarium. So the convention of referring to lifestyle risk factors as ‘modifiable’ risk factors is misleading as it implies we know how to achieve behaviour change and that there are highly effective interventions ready to be operationalised. Perhaps ‘theoretically modifiable’ or ‘potentially modifiable’ would be more realistic terminology?
Within the last 4 years the National Institute for Health and Clinical Excellence (NICE) has published guidance on tackling smoking cessation,3 increasing physical activity,4 preventing harmful drinking,5 and the prevention and management of obesity.6 In a recent edition of the BJGP Mercer gave a balanced review of the NICE clinical guidelines on obesity for general practice.7 He highlights deficiencies in the data underpinning the recommendations: little of the evidence is derived from studies focused on a primary care setting, conducted in the UK, or involving patients rather than volunteers. To a greater or lesser degree, Mercer's observations are generalisable to all the NICE guidelines on lifestyle modification.
Further evidence is required to address the deficiencies highlighted above, and future research also needs to address those issues that arise from the complexities of day-to-day general practice. For example, in a consultation with an already overcrowded agenda, how does one most effectively introduce discussion about lifestyle and behaviour change without disenfranchising the patient? What are the best approaches with patients who have repeatedly been unable to sustain lifestyle change? How can the primary care clinician encourage lifestyle change for the patient who already is challenged by coping with everyday life and has severe constraints on their money and time?
Considering the conclusions of Iversen and colleagues' study in the context of real patients, competing agendas, and paucity of evidence, one might easily get despondent despite a strong desire to achieve the best for one's patients.
McLandburgh Wilson said ‘Twixt the optimist and pessimist the difference is droll: the optimist sees the doughnut, but the pessimist sees the hole’.8 We have a professional responsibility to remain optimistic, while seeking robust evidence to inform our primary healthcare practice and our commissioning of interventions to modify patients' lifestyles.
Notes
Provenance
Commissioned; not peer reviewed.
- © British Journal of General Practice, 2010.