Recent position statements have re-affirmed the benefits of an active lifestyle [1, 2]. The current physical activity recommendation for adults, aged between 18–65 years, to promote and maintain health is to accumulate at least 30 minutes of moderately intense physical activity on at least five days of the week. Promoting accumulative, lifestyle physical activity is an ideal approach to combat the high levels of inactivity evident in global populations [3, 4]. Brisk walking has been suggested as the mode of physical activity most likely to increase physical activity at a population level [5] and is the most commonly reported mode of physical activity amongst adults in many populations [3, 6]. It is available to almost all individuals with little risk of injury, is a no cost activity and it can be incorporated into peoples' daily routines [7]. Researchers have identified that self determined brisk walking, even in short bouts of 10 minutes, for 30 minutes a day (including simple everyday walking activities such as walking a dog) produce moderate physical activity at the intensity required to achieve health benefits [8, 9].

Walking interventions can be effective in reducing body weight, body mass index (BMI), waist and hip circumference, body fat, blood pressure and the cholesterol:high-density lipoprotein (HDL) ratio [10–16] and may be effective in improving mood, affect [14, 17, 18] and quality of life [19]. Conversely, some studies have demonstrated that a walking intervention is not sufficient to affect any of these health-related outcomes [20–24]. The reasons for such equivocal results are unclear, therefore determining the potential health benefits that can be achieved through walking is crucial to the public health message.

Whilst several meta-analytical and systematic reviews exist that examine how best to promote physical activity [25, 26] there is comparatively limited evidence on the most effective methods to specifically promote walking. A recent systematic review from Ogilvie and colleagues (2007) examined the effectiveness of interventions aimed at increasing walking at both an individual and population level. The review concluded that the strongest evidence exists for tailored interventions that are targeted at individuals most motivated to change. The authors suggested that future studies should also attempt to examine whether walking interventions "are sufficiently frequent, intense, or sustained to produce measurable outcomes in anthropometric, physiological, biochemical or clinical outcomes" (Ogilvie et al., 2007 p.1207).

One category of intervention, discussed by Ogilvie and colleagues, was the use of pedometers as an integral component of the intervention. A substantial body of research exists that supports the use of pedometers to measure physical activity [27, 28]. A recent systematic review examined the association between pedometer use, physical activity levels and a variety of health related outcomes [29]. The authors concluded that pedometer use was significantly associated with increased physical activity levels and reductions in BMI and systolic blood pressure. In 2006 the National Institute for Health and Clinical Excellence (NICE) in the United Kingdom produced a review of pedometer-based intervention studies between 1990 and 2005 [30]. Due to stringent inclusion criteria, conclusions from this review were drawn from only four studies. Both reviews provide support for the suggestion that pedometers may be useful motivational tools for increasing walking. However, there are several limitations when considering the volume of published studies in this area highlighted by these reviews. Studies were predominantly of short duration (< 12 weeks) and based in the United States of America (USA) with small samples consisting mostly of clinical sub-populations. There is limited evidence regarding their effectiveness in non-clinical samples or in countries other than the USA. Additionally, few studies reported more than one outcome variable of interest. There is a need for cross-cultural, sufficiently powered randomized controlled trials to further examine the effectiveness of pedometers in a community setting.

Evidence from two of the studies included in the NICE review suggests that a support structure, that addresses social and cognitive factors, is required for a pedometer-based intervention to be effective [22, 31]. A physical activity consultation, using a theoretically grounded framework, constitutes one method of addressing these factors that has been demonstrated to effectively promote physical activity [26, 32]. Researchers have suggested that pedometers may provide an important point of discussion for in-depth physical activity consultations [33] yet this is an area that has not been particularly well addressed by pedometer-based intervention studies.

The Walking for Well-being in the West (WWW) study is a multi-dimensional community-based randomized controlled trial which aims to promote and maintain increased walking behavior through the use of physical activity consultations and a pedometer-based walking program. The purpose of this article is to investigate the short term effects of a pedometer based intervention, in conjunction with a physical activity consultation, on walking behavior and health related outcomes in individuals not meeting current physical activity guidelines. Full details of the study design and rationale are reported elsewhere [34] Details of the subsequent stages of the WWW intervention study can be found on the Scottish Physical Activity Research Collaboration website http://www.sparcoll.org.uk.

This study is one of the first adequately powered, UK based randomized controlled trials to examine the impact of a pedometer-based walking intervention on step-counts in a community setting. The major finding of this study was that a graduated pedometer-based walking program, in conjunction with a physical activity consultation increased walking in low-active adults over a period of 12 weeks. The control group, included to account for the intrinsic motivation of volunteer participants [63], displayed no significant change in steps/day over time. The conservative intention to treat analysis (baseline carried forward for missing values) produced a mean change in the intervention group of 3,175 steps/day, an increase of 47% above baseline values, a favorable increase compared with other pedometer-based randomized controlled trials [22, 31, 33]. In their systematic review, Bravata and colleagues (2007) reported that pedometer users increase physical activity by an average of 26.9% over baseline values.

The overall goal for participants was to increase their baseline step-counts by 3,000 steps/day on five days of the week (i.e., equivalent to current physical activity recommendations); equating to an overall weekly increase of 15,000 steps/week. The intervention group displayed a mean increase in weekly step counts of 22,225 steps/week thus exceeding the recommended goal. At an individual level 64% of participants in the intervention group achieved the goal of 15,000 steps/week. Participants in the intervention group progressed from being classified as "low active" (5,000–7,499 steps/day) to within 23 steps of being classed as "active" (≥ 10,000) according to suggested public health ranges for pedometer counts [64]. These results suggest that the intervention is a successful method of promoting walking and allowing individuals to meet suggested public health recommendations.

One issue of the study, also discussed in the Bravata et al. systematic review, is that it is not possible to disentangle the respective contributions of the physical activity consultation from the benefits of the pedometer-based goals of the walking program. Physical activity consultations have been demonstrated to be effective at promoting physical activity [32, 54], and this particular walking program has previously been shown to be effective at promoting short-term increases in walking [58]. Consistent with recommendations from Bravata and colleagues, the next stage of the WWW project involves the control group following the 12 week walking program without the additional physical activity consultation, while the intervention group will be provided with continued support in the form of a further physical activity consultation. These longer term comparative results will help to determine the most effective components of this intervention.

The increase observed in the step-count data was supported by the self-reported results of the IPAQ; an increase in reported minutes of leisure time walking was found in the intervention group (median increase of 100 minutes per/week). While it is not directly comparable, the increase of 22,225 steps/week observed in the intervention group is approximately equivalent to an additional 222 minutes of walking per week. Although we must consider possible errors in perception when recalling walking [65] this suggests that the primary means by which participants increased their walking was during their leisure time, consistent with previous research [66]. The discrepancy between the objective and subjective measures of activity could be partially explained by the pedometer measuring all activity and the IPAQ only measuring activity in bouts of 10 minutes and greater. Alternatively, additional increases in step-counts could be attributed to an accumulation of walking in other domains and other non-significant increases in ambulatory activity. This may be demonstrated in the significant difference in occupational and total minutes of walking recalled between the intervention and control groups at week 12.

The IPAQ also revealed that the intervention group reduced time spent sitting during both weekdays and weekend days. Researchers have suggested that sedentary behavior, such as time spent sitting, is positively associated with coronary heart disease risk factors, obesity and development of the metabolic syndrome [67–69]. It has been suggested that interventions should seek to both increase physical activity and to decrease sedentary behavior [70]. In this study we did not aim to directly reduce sedentary behavior and as the IPAQ does not measure sitting time across a specific domain it is impossible to determine where the reductions in sitting time occurred. Future interventions incorporating physical activity consultations should ensure that identifying and reducing sedentary behaviors is an integral part of the consultation process and that any physical activity questionnaires used should describe sedentary behaviors to the same extent as physical activity behaviors. These results should be treated with a degree of caution given the disparity between the reduced time spent sitting as indicated by the IPAQ (4 hours of total sitting per day) and the increase in physical activity as measured by the pedometer (approximately 30 minutes of activity per day). It may be possible that non-sitting activities such as standing may attribute to a proportion of this difference however this does not provide an adequate explanation given the magnitude of this observed discrepancy. A more probable explanation is that the error and limitations associated with self-report measures of physical activity [71] are also associated with self-report measures of sedentary behaviors. Preliminary evidence that compares activity levels as measured by a pedometer and the IPAQ would suggest that direct comparisons are unwise as evident by low correlations between the two methods [72].

Outcome measures were only assessed at baseline and week 12. Participants were asked to record their pedometer readings from intermittent sub-goals in self-recorded diaries. Completion of these was not an essential requirement of participants and as such these were often incomplete. The level of missing data in these diaries made it impractical to perform statistical analysis on a week by week basis. We were reluctant to constantly monitor participants to try and mirror a 'real-world' scenario. As a result we were unable to determine whether participants actually achieved, or exceeded, their targets in the periods between assessments. Therefore, the possibility remains that, in a worst case scenario, participants increased their physical activity levels only during week 12 in order to achieve their final target; although such a substantial increase in step-counts over a one week period would be unlikely. A previous study from our research group [58] that followed individuals throughout the intervention reported an incremental increase in step-counts on a week by week basis. While only measuring at pre and post intervention may be seen as a limitation of the current study, this approach removes the possibility of increased motivation through researcher presence and may also decrease the risk of participant drop-out due to repeated meetings. Making completion of step-count diaries a requirement of participants would have increased the level of data available for analysis. However, this substantially increases the responsibility on the participant which, over a prolonged time-period, may provide a potential source for participant drop-out. Conversely, filling in the step-count diaries on a regular basis may act as an extrinsic motivational factor; which could potentially mask the effects of the walking program or the consultation. The recent systematic review on pedometer use suggests that future trials compare pedometer use with versus without a step-count diary [29] and the results of this study support this suggestion given the issues discussed above.

In this study contact with intervention group participants was minimal and yet significant behavior change was still achieved. Following the RE-AIM framework for health behavior interventions [73], this intervention was simple to implement, was efficacious at an individual level and has the capacity to be adopted and implemented within a variety of real-world settings.

Whilst the current study was successful in increasing the daily walking of participants, there was no effect on any of the health-related outcomes measured other than a small to medium, significant increase in positive affect reported by the intervention group. There is contradicting evidence regarding the relationship between walking and affect. Some researchers have suggested that walking may not be performed at a sufficient intensity to produce corresponding positive changes in affect [74]. Conversely, the current study supports research that has demonstrated a positive benefit on affect following a walking intervention [14, 18, 75]. Researchers have also demonstrated that short, acute bouts of walking may also improve affect [17, 76]. These findings provide support for walking as a pleasurable activity and it has been proposed that this enjoyment is linked to intrinsic motivation and subsequent adherence to physical activity [76].

The lack of changes in the other health-related outcomes measured in this study may not be unexpected given the pragmatic approach of the walking intervention. The walking program was designed to increase participants' walking in graduated bi-monthly stages: participants began the program by accumulating an additional 1,500 steps on at least three days of the week for the first two weeks before progressing with the frequency of this goal and then the quantity of walking. This approach was utilized to allow participants the opportunity to follow progressive short term goals in order to reinforce successful strategies or attempt alternative strategies in order to achieve their goals. Subsequently, if this approach was followed by participants, the goal of accumulating an additional 3,000 steps at least five days of the week was only applicable for the final six weeks of the intervention. Recent meta-analyses [13, 15], that have found decreases in body weight, BMI, percentage body fat, blood pressure and the cholesterol:HDL ratio after a walking intervention, report means and ranges of duration and frequency consistent with the 30 minutes (3000 steps), five days a week target promoted in the current study. These meta-analyses used data from studies ranging between eight and 104 weeks in length.

It may be the case therefore, that a period longer than six weeks at this duration and frequency is required to allow positive physiological changes to occur. Future follow-up stages of the WWW study will attempt to evaluate this aspect. However, as previously mentioned; we are unable to conclude whether participants strictly followed the graduated approach or whether they were successful in doing so throughout the program. Indeed, previous work investigating the dose-response relationship has shown that it can take up to two years for an increase in HDL to occur as the result of exercise training [77]. On the other hand similar changes in HDL and other anthropometric variables have occurred after 12 weeks in other investigations [16, 78]. It is possible, therefore, that the current intervention may have resulted in positive health benefits if it had continued.

Although we may not have expected physiological changes at week 12 due to the graduated walking program, we must consider alternative methodological considerations of the study that may have contributed to the lack of significant changes in health outcomes. With the exception of BMI, participants' health outcomes values were deemed to be within normal ranges at baseline. Significant, and clinically meaningful, changes will consequently be harder to achieve in this population and may indeed be unnecessary for many participants. It has been recommended that 45–60 minutes of moderate physical activity per day is required for weight maintenance [79]. Therefore it was unlikely that these changes would have occurred following an increase of approximately 30 minutes of walking. This study also did not attempt to control for dietary factors. Previous studies have demonstrated that calorific restriction is a more effective weight loss tool than physical activity [80]. As we did not control for or monitor diet it is possible that participants may have compensated for the increase in physical activity by increasing energy intake or by deceasing energy expenditure during other parts of the day. Such a possibility is supported by previous research showing that there is an increase in appetite after a single bout of exercise [81] although this is not always been the case [82]. In the current study no measure of energy intake/expenditure was made hence we are unable to determine whether an increase in energy intake, or decreasing energy expenditure in other parts of the day, has counteracted the increase in energy expenditure due to the walking.

Finally, it may be that the intensity of the walking undertaken by the participants was not sufficient to stimulate health benefits. It is possible that the changes in step-count observed in our participants were caused by short (less than 10 minutes) bouts at low speeds. To date there has been little community based work where the intensity of physical activity has been closely monitored. Results from this and similar community based studies [14, 16, 23] present equivocal and conflicting findings of producing corresponding physiological changes following successful behavior change. This illustrates that in comparison with controlled, laboratory based studies positive changes in health may be more difficult to achieve in a real world setting, where the frequency, duration and intensity of the intervention cannot be objectively measured.

In summary, this study has demonstrated that a 12-week pedometer-based walking program in combination with a physical activity consultation was an effective way to increase walking, reduce sedentary behavior and increase positive affect in a community based sample not meeting current physical activity recommendations. The intervention was relatively simple to implement and has the capacity to be reproduced in a variety of settings. The intervention was not sufficient to induce beneficial physiological changes. Future stages of this study will examine adherence to the intervention, compare the effects of the walking program with versus without the physical activity consultation and provide longitudinal data on health related outcomes.