Brief lifestyle advice has been defined as advice delivered in a single consultation (Bull and Jamrozik, 1998), often opportunistically, that is usually of only 2–3 minutes’ duration (Lewis and Lynch, 1993). It is commonly delivered in primary care and supported by written information (Calfas et al, 1996; Ashenden et al, 1997; Kreuter et al, 2000). Brief lifestyle advice often addresses a variety of modifiable behaviours including diet, physical activity, alcohol consumption and smoking (Ashenden et al, 1997). Interest in brief lifestyle advice has been increasing since it was identified that modifiable lifestyle factors including diet (Liu et al, 2000; Boekholdt et al, 2006; Myint et al, 2008), smoking (Gordon et al, 1974), alcohol consumption (Foerster et al, 2009) and physical activity (Hakim et al, 1998; Manson et al, 2002; Barengo et al, 2004; Hu et al, 2004) are associated with risk of future disease. Type 2 diabetes and obesity, both of which have been described as “epidemics”, provide excellent examples of diseases in which early, effective lifestyle interventions could prove lifesaving. The NHS Future Forum has recently recommended that all NHS healthcare professionals make “every contact count”, a proposal that encourages lifestyle advice delivered at every contact with an NHS professional (Bailey et al, 2011). This is to be supported by the UK government (Lansley, 2012). In addition, there are already financial incentives in place to provide lifestyle advice in primary care with the aim of reducing overall risk in the population (British Medical Association and NHS Employers, 2009). However, concerns have been raised that public funds may be directed towards interventions for which there is little robust evidence of success (Ashenden et al, 1997).
A systematic review published in 1997 of 37 trials investigating the impact of brief and more intensive lifestyle advice versus no advice on physical activity, diet, smoking and alcohol consumption found that while the results were in general, promising, there was insufficient evidence to conclude that lifestyle interventions could produce substantial behaviour change (Ashenden et al, 1997). Most of the trials that were included followed participants for at least 1 year. This systematic review found that the odds ratio for smoking cessation following smoking cessation advice versus no advice was 1.32 (95% confidence interval [CI] 1.18–1.48) and that more intensive versus brief advice had no impact on the success of the intervention (odds ratio 1.07 [95% CI 0.88–1.29]). For diet, alcohol consumption and physical activity, the trials were too heterogeneous to be combined in this way.
The purpose of this review was to examine the more recent evidence to determine the effectiveness of the provision of brief lifestyle advice in primary care.
Method
Search strategy
A search of the MEDLINE database was conducted including articles published between 1996 and January 2012. Studies from this period were included to update the findings of the 1997 systematic review. A specific list of search terms were used to identify relevant studies (Box 1).
In total, 469 potentially relevant articles (including reviews of the literature) were identified through this search strategy. The abstracts for studies were retrieved and the reference lists from review articles were searched. From this initial survey a set of inclusion criteria was developed and the final selection of trials to be included was made. A total of 19 trials were selected.
Inclusion criteria
The review was restricted to studies published in the English language that investigated the provision of brief lifestyle advice to adults in primary care. Only studies with a randomised design were included. The authors examined studies that provided advice concerning physical activity, diet, alcohol consumption or tobacco smoking. Studies that included advice on more than one of these behaviours were included. Studies were included if brief advice was compared with either no advice or usual care or with a more intensive intervention.
Brief advice was defined as being of less than 30 minutes’ duration, delivered face-to-face in a single consultation with no further follow-up or reinforcement by a person routinely employed in primary care (GP or practice nurse). In general, advice was provided verbally and supported by written materials. Trials were included irrespective of the duration of follow-up but needed to include a measure of behaviour change (for example, change in body weight, blood pressure or physical activity questionnaire score). Studies were excluded if the only outcome measure was thoughts about behaviour change or change in stage of change. Trials were also excluded if the advice was accompanied by a pharmacological agent (for example, nicotine gum or patches). The major reasons for excluding trials were:
- Not relevant to review (trial protocols, cost-effectiveness analyses, duplicate studies, cross-sectional studies of advice given in general practice).
- Non-randomised study design used.
- No brief advice intervention arm.
- Review articles.
Data extraction
The authors extracted data from the published studies.
Quality assessment
Studies were assessed using the simplified scheme described in the 1997 systematic review described earlier (Ashenden et al, 1997; Enkin et al, 2000). Three dimensions of trial methods were assessed: selection bias at entry to the study (for example, was a robust method of randomisation used?), selection bias after trial entry (for example, were analyses by intention to treat and were rates of follow-up high and similar in all groups?) and selection bias in assessing outcomes (for example, were individuals collecting and inputting data blinded to an individual’s intervention allocation?). A score of 3 was awarded if maximal attempts had been made to control bias and 1 if little or no attempt had been made. In some instances the reporting of the study was such that a low score had to be awarded if no information was provided to the contrary.
Data analysis
The trials included reported many different outcomes in a variety of ways. In view of this, data were extracted, summarised and tabulated. Outcome measures included self-reported changes in smoking behaviour, physical activity and diet, biochemical measures of smoking cessation including salivary cotinine and objective measures of fitness including a 6-minute walk test.
Results
Characteristics of trials
A total of 19 trials were included in the review (Table 1). Most physical activity trials recruited sedentary adults and most smoking cessation trials recruited adult smokers irrespective of other health conditions or comorbidities. A minority of studies recruited healthy adults. The studies included a broad range of ages from 15 to 80 years old. All studies included both male and female participants. The studies included a mixture of those that compared brief lifestyle advice with no advice or usual care and those that compared brief lifestyle advice with more intensive interventions. In some instances a variety of approaches were tested in the same trial.
In the majority of trials the advice was provided by a GP, with fewer studies choosing a nurse to deliver the intervention. Most advice was supplemented by written materials.
Follow-up varied between 1 and 24 months, with the majority collecting data for between 6 and 12 months.
The quality of the studies included was generally good. The most common reason for being awarded a low score was selection bias after trial entry, with many studies reporting high losses to follow-up.
Effectiveness of brief lifestyle advice
A summary of the results of the 19 studies included in this review is shown in Table 2. Overall, in the 10 studies that compared brief lifestyle advice with more intensive advice, three studies demonstrated that brief advice could be as effective as more intensive advice, six studies showed that brief advice was not as effective as more intensive advice and one study produced an equivocal result. In the 10 studies that examined the provision of brief lifestyle advice compared with the provision of usual care or no advice, six showed a positive result for the effectiveness of brief lifestyle advice, one showed a negative result and three showed equivocal results. (Please note that one of the studies [Harland et al, 1999] had brief, intensive and control arms and was therefore considered in both the groups of 10 studies above.) These results need to be qualified against the fact that in the vast majority of the studies included, outcome measures relied on self-reported behaviours. In addition, in many of the studies, follow-up of participants was not ideal with few studies being able to collect data on more than 80% of participants at follow-up. There were also too few studies that examined the impact of the provision of advice concerning more than one behaviour to draw any useful conclusions concerning this approach.
Discussion
The evidence for the effectiveness of brief intervention in primary care remains unclear. Several studies suggest that there may be some health benefits associated with the provision of brief lifestyle advice and that brief lifestyle advice may be nearly as effective as more intensive regimens, particularly in the short term. The findings of this review are broadly in keeping with those of the UK Health Development Agency. In its evidence briefing for brief interventions aimed at increasing physical activity levels (Hillsdon et al, 2005) the agency concludes that at the present time the evidence suggests that a single episode of brief tailored advice with some follow-up can increase physical activity, at least in the short term (Hillsdon et al, 2005). This briefing goes on to highlight various components that were associated with the greatest likelihood of success. These included the provision of tailored information delivered verbally with written support, setting goals, self-monitoring, exploring beliefs about physical activity, ongoing verbal support, providing occasional reviews, promoting moderate-intensity activity such as walking and no requirement for attending a facility (Hillsdon et al, 2005).
The evidence that brief lifestyle advice may be no less effective than more intensive advice is encouraging for more widespread adoption in routine clinical practice, particularly in primary care. The time pressures on primary care practitioners are increasingly being recognised and there now seems to be an acknowledgement that studies attempting to deliver lifestyle advice in primary care need to develop interventions that can be delivered in just a few minutes (Screening and Intervention Programme for Sensible drinking [SIPS], 2008). In addition, more robust studies are being conducted to further evaluate brief lifestyle advice in comparison with a more intensive intervention.
One of the major shortcomings of many of the studies reviewed is the lack of objective measures of behaviour, with many relying on self-reported information. There is a clear need for robust objective outcome measures of behaviour to be included in randomised trials of lifestyle interventions and not just intermediate measures (for example, intention to change behaviour and thoughts about the need to change behaviour), as have often been used in the past (Ashenden et al, 1997). Physical activity can be measured objectively using pedometers or accelerometers. Serum cotinine (Jarvis et al, 2003) is a reliable indicator of tobacco smoked and plasma vitamin C provides a robust estimate of fruit and vegetable intake (Khaw et al, 2001).
One of the authors recently carried out a pilot randomised controlled trial analysing the effect of brief intervention as well as communicating personalised cardiovascular health risk on physical activity (Price et al, 2011). This trial followed 194 adults with an increased risk of cardiovascular disease. Participants were randomised to receive an estimation of their 10-year cardiovascular risk and then randomised again to receive a 15-minute brief intervention reinforced with written materials. The primary endpoint was activity measured with an accelerometer. Further endpoints included weight, blood pressure, high-density lipoprotein cholesterol, serum cotinine in smokers, triglyceride levels and further estimated 10-year risk.
There were no significant differences between individuals allocated to receive brief intervention versus those receiving no intervention in most primary or secondary end points. Men had a 2.6% decreased waist circumference (P=0.006) but there was no difference in women. There was a 7.9% reduction in serum cotinine in smokers (P=0.028) and a 10.2% decrease in triglycerides (P=0.02; Price et al, 2011).
The methods used to study lifestyle interventions need to adapt in order to increase the evidence base for the success of lifestyle interventions. Many studies were excluded from this review because they did not use a randomised design.
The 1997 systematic review of lifestyle advice published by Ashenden and colleagues highlighted many of these issues and the inability to draw firm conclusions from the studies conducted up until then (Ashenden et al, 1997). Disappointingly, heterogeneous outcome measures, outcomes measured by self-report and non-randomised trial designs still plague this area of research. Future studies need to address these issues in their trial design.
It is possible that over recent years the need to provide lifestyle advice may have diminished because of better knowledge among the population about healthy lifestyle choices. However, despite the Government’s focus on improving the health of the population in recent years, the evidence suggests that public perceptions of what constitutes a healthy lifestyle and which behaviours are unhealthy remain poor. In one UK survey of 512 adults aged 17–45 years, 25% of responders were current smokers yet only 4% recalled being advised to quit smoking (Duaso and Cheung, 2002). When asked what were their main sources of health promotion information, 67% of study participants cited magazines and 47% television. In younger people the Internet appears to be an important source of lifestyle advice (Humphreys and Costarelli, 2008).
In another study of 390 adults with and 190 without cardiovascular disease (CVD) risk factors that aimed to determine which factors predict recall of lifestyle advice, those with CVD risk factors, men and older participants were the most likely to recall that they had been given lifestyle advice (Little et al, 1999). Those able to recall lifestyle advice were more likely to report a healthier current lifestyle than those who could not recall lifestyle advice. It is of concern that 30–50% of responders who reported an unhealthy lifestyle were unaware that it was unhealthy. This study included adults from a wide range of socioeconomic backgrounds. Even in high-risk individuals, recollection of lifestyle advice can be poor. In a study of over 3000 people who had been hospitalised with heart failure and received lifestyle advice during their admission, only 46% of them could recall receiving the lifestyle advice (Lainscak et al, 2007).
These studies confirm that while much effort has been expended on increasing the health of the nation and increasing awareness of what constitutes a healthy lifestyle, there remains a significant proportion of the population who are unaware of rudimentary health messages (the benefits of not smoking, participating in physical activity, eating a low-fat, high-fibre diet containing plenty of fruit and vegetables and consuming alcohol in moderation). In view of these findings, it is important that research continues to be conducted to investigate approaches to educating the population about healthy behaviours.
Conclusion
As diseases such as obesity and type 2 diabetes continue to increase in prevalence they present a growing challenge to 21st Century healthcare professionals. The pathophysiology and therapeutic options for these conditions are increasingly well understood. Modifiable lifestyle factors such as smoking, diet, alcohol and exercise would appear to be more straightforward targets for intervention. Despite the shortcomings of the available data as one of the studies concludes (Lancaster et al, 1999), brief lifestyle advice may result in some benefits in particular groups of people and rarely results in adverse outcomes. It can be used as a useful adjunct in clinical care. It is, however, clear that insufficient data exist to champion this method at the expense of other more evidence-based interventions.
