Evaluating the effectiveness of implementation intentions in a pedometer worksite intervention

Abstract

BACKGROUND:

Pedometer-based worksite interventions have been found to be successful in increasing physical activity (PA) but adherence is challenging.

OBJECTIVE:

To examine the use of Implementation Intentions (II), a self-regulatory skill, with self-monitoring with a pedometer to initiate behavior change as well as post-intervention adherence in a worksite wellness intervention.

METHODS:

University employees (N = 54) participated in an 8-week pedometer-based intervention. A 2-arm randomized trial was used to compare the effectiveness of 1) only pedometers (PED) (n = 28) and 2) pedometers and II (PED+II) (n = 26) on PA.

RESULTS:

Significant differences were observed between time points (p < .0001) but not between groups. Post-hoc pairwise comparisons between the time points revealed difference between Baseline and Week 4 (mean difference: 2446.9 steps/ day; p < 0.001), Week 4 and 12 (mean difference: 2956.3 steps/ day; p < 0.001), and Week 8 and 12 (mean difference: 2228.8 steps/ day; p = 0.005).

CONCLUSION:

The PED+II group had higher step increases during the intervention indicating that the behavioral strategy was effective. However, participants in both groups had a significant decrease in steps from the end of the intervention to the delayed-post assessment highlighting the challenge to maintain behavioral changes post-intervention.

Keywords

Pedometry physical activity health behavior intervention

1 Introduction

US adults spend the majority of their waking hours at their worksites [1] especially as society has shifted from a traditional 8-hour workday to a longer workday [2]. Worksites serve as a community with existing social support, procedures for policy and environmental changes, and cultural change [2]. Approximately 60% of US adults are employed [3], demonstrating that workplaces show promise for health promotion programs. Employers should invest in their employees’ health due to increasing health care costs [4] and work productivity losses [5]. As such, worksites show promise as an avenue for health promotion programs.

Currently, only 48% of American adults meet physical activity and public health aerobic activity guidelines of at least 150 minutes of moderate-intensity, aerobic physical activity each week [6]. Not only are the majority of Americans not meeting physical activity guidelines, they are becoming increasingly sedentary during their work day [7]. Since 1950, sedentary jobs have increased by 83% with fewer than 20% of jobs being physically active compared to 50% of an active workforce in 1960 [7]. Since the health detriments of physical inactivity and sedentary behavior are well-documented [8 –10], it is important to examine strategies that can increase physical activity and decrease sedentary behavior.

It is of public health importance to promote physical activity at the worksite due to the decrease in leisure-time physical activity coupled with the increase in workplace sedentary behavior and time spent at the workplace. Traditional worksite-based physical activity programs have focused on physical activity adoption rather than physical activity adoption and adherence. Since 50% of adults beginning a physical activity program quit within the first few months [11], it is critical to examine behavioral strategies that may enhance adherence to physical activity.

Often, participants lack the self-regulation skills necessary to overcome obstacles encountered in sustained behavioral change [12]. Behavior change is most likely realized when the individual has a strong intention to perform the behavior (i.e., goal or behavioral intentions) and has developed volitional strategies (i.e. implementation intentions) for behavioral performance [13]. Implementation intentions are if-then plans that connect good opportunities to act with responses that will be effective in accomplishing one’s goals [12]. These are specific plans of action concerning when, how, and where an intended behavior will be enacted [12, 13]. For example, if an individual walks outside to achieve activity goals, then an implementation intention plan of action might be, “If it is raining outside, then I will walk on the track in the rec.” In respect to PA, implementation intentions have had positive effects on PA behaviors [13]. Implementation intentions are a self-regulating strategy that individuals can utilize to identify potential behavior obstacles and initiate the desired goal-directed response.

Lifestyle focused PA interventions often utilize body worn technologies that support behavior change. Pedometers are step counting tools that provide feedback and enhance self-monitoring. Recent meta-analyses have found that pedometer interventions have a moderate and positive effect on the overall increases of PA [14] and increase baseline walking levels by over 25% [15]. Pedometer-based interventions help increase daily physical activity but it is unknown to what extent observed changes are sustained or if it is possible to accrue benefits over long-term adherence [16].

While behavior strategies such as implementation intentions may be critical to adopting and adhering to physical activity, physical activity monitoring devices may also play a role in adoption and adherence. Although use of pedometers has a positive impact on increasing physical activity [14, 17] it is still unknown whether pedometer usage increases physical activity adherence. Furthermore, using pedometers and implementation intentions simultaneously in the worksite setting has yet to be studied. As such, the purpose of this study was to examine the use of implementation intentions, a self-regulatory skill, with self-monitoring through the use of a pedometer to initiate behavior change as well as post-intervention adherence in a worksite wellness intervention.

2 Methods

2.1 Participants

In the fall of 2015, faculty and staff at a medium sized, Midwestern university received a recruitment email to participate in a walking program. Inclusion criteria included: 1) between the ages of 18–64 years, 2) current full-time employee at the university, 3) pass a Physical Activity Readiness Questionnaire (PAR-Q), and 4) currently not meeting physical activity (PA) guidelines of 150 minutes of moderate or vigorous PA/week. Sixty-eight participants enrolled in the program with 56 participants completing the intervention, resulting in an 82.4% retention rate. Participants were required to have four or more days of step count data each week to be included [16]. Two participants were excluded from data analysis due to incomplete data resulting in a sample of 54 participants. The Institutional Review Board Human Subjects Research Committee approved the study in 2018. Participation was completely voluntary and written informed consent was obtained by all subjects.

2.2 Design

The intervention was an 8-week, randomized control design. After participants were enrolled in the program, they were randomly assignment to one of two groups: 1) the Pedometer Only Group (PED) (n = 28), or 2) Pedometer+Implementation Intentions Group (PED+II) (n = 26). The intervention was broken into three phases: Phase 1: Participant Engagement Stage (Weeks 0–4), Phase 2: Participant Autonomous Stage (Weeks 5–8), and Phase 3: Adherence Stage (Week 12). In order to isolate the influence of implementation intentions on physical activity, participants in both the PED Only and PED+II groups engaged in all the same activities throughout the study with the exception that the PED+II group completed implementation intentions during Weeks 0 and 4. See 2 Fig. 1 for intervention timeline.

Fig.1

Intervention timeline.

Baseline: There was an Orientation Meeting in Week 0 in which participants were assigned a blinded pedometer to wear each day to record steps, however; the pedometers were modified so that participants were unable to view the data during this phase. By blinding the pedometers, participants were not able to use the step data to influence their activity and thus, get a more true representation of their current activity levels. Participants were then educated on how to properly wear the pedometer. Height and weight measurements were taken.

Phase 1: Participant Engagement Stage (Weeks 1–4): In Week 1 the pedometers were unblinded so that participants could now view the step data in real time as they wore the pedometers. The participants were then taught how to use the pedometer and manually record data in their step logs daily. Participants were asked to wear the pedometers every day for the duration of the intervention and record their data. Weekly behavioral strategy modification meetings were held during Weeks 1–4 of the program. The behavioral modification activities used in the meetings were modified from the Diabetes Prevention Program’s Lifestyle Change Program [18]. Participants were taught the behavioral strategies goal setting and modification, identifying motivators as well as barriers, and relapse-prevention. Information regarding current physical activity recommendations and campus resources to engage in physical activity was also provided. Meeting structure started with an informational component followed by an activity that engaged the participants in the behavioral strategy being taught. For example, in the meeting that examined motivating factors, participants were first taught what motivation is and different types of motivation (i.e. intrinsic and extrinsic). Intrinsic motivation was described as engaging in behaviors for their own enjoyment or interest and for not any other separable outcome [19]. Extrinsic motivation was explained as engaging in behaviors for some separable outcome such as weight loss or health benefits [19]. Then participants engaged in an activity in which they listed all their motivators for being physically active. Participations were asked to then rank their top three motivators and reflect on how they could use their motives to remain active. Participants were encouraged to attend meetings in order to be eligible for a drawing at the end of the intervention. At the end of Week 4, participants completed a mid-point online survey.

Phase 2: Participant Autonomous Stage (Weeks 5–8): Participants were asked to continue to wear their pedometers daily and log their steps for the remainder of the intervention. The weekly meetings were discontinued and participants were independent during the Autonomous Stage. However, weekly email reminders were sent to participants to log their steps and meet their step goals. At the end of the beginning of Week 9 (after eight full weeks of data collection), participants turned their pedometers and step logs in. Height and weight measurements were taken again.

Phase 3: Adherence Stage (Week 12): Participants were reissued a blinded pedometer for a delayed post-assessment four weeks after the intervention ended. Thirty-six participants (64%) completed the delayed post assessment. At the end of Week 12, the pedometers were collected and step counts recorded.

Online Surveys: Participants were emailed links to an online survey during Weeks 0, 4, 9, and 12. The online survey collected data on participants’ demographics, self-reported physical activity, and implementation intentions for participants in the PED+II group.

2.3 Study variables

Baseline Steps: Participants were assigned blinded pedometers to wear for one week (Week 0) for baseline averages. The blinded pedometer had a zip tie around it so that the participant could not view their daily step counts. Participants were instructed to wear the pedometer daily and not to alter their normal activity behaviors. The pedometers have a seven-day memory recall, so at the end of Week 0, participants cut the zip ties and were able to get their step counts for the previous seven days.

Daily Steps: Participants were each assigned a New Lifestyles NL-1000 Accelerometer for the duration of the intervention. When the pedometers were assigned, instructions on how to wear and use the pedometer was reviewed and discussed. Participants were asked to put the pedometer on first thing in the morning and take it off right before bed at night. They were then asked to record their steps daily in the step logs. If they forgot one day, the seven-day recall on the pedometer could be used to retrieve the data.

Participants were required to have four or more days of step count data each week to be included. This is based on the results from Tudor-Locke et al. [16] that four or more days of pedometer monitoring provide a reliability of intra-class correlations greater than 0.90 for predicting weekly physical activity of adults.

Body Mass Index: At Baseline (Week 0) and the conclusion of the intervention (Week 9) height was measured on a stadiometer and weight was taken on a Tanita scale. Body mass index (BMI) was thencalculated.

Implementation Intentions: Implementation intentions are specific plans of action concerning when, how, and where an intended behavior will be enacted [13]. The plans are formatted into if-then statements. For example, “if situation Y occurs, then I will initiate goal-directed behavior Z” [12]. Participants in the PED+II group completed these implementation intentions online as part of their study surveys at Baseline and Week 4.

The participants were prompted with the following message modified from Stadler, Oettingen, and Gollwitzer [20], “Many situations may place you at risk for quitting your physical activity plan. First, identify three obstacles you perceive to meeting your daily step goal.” Then participants listed three obstacles. Next participants were prompted with the following message, “Then, make your plan with how you would deal with these situations. For example, an obstacle identified could be sleeping in late and not being able to go for a walk in the morning. The plan to deal with this obstacle would be: If I get up too late, then I will go for a walk over lunch.” Participants then completed the following statements: If _ _ _ _ _ _ _ _ _ _ _ (obstacle #1), then I will _ _ _ _ _ _ _ _ _ _ _ (plan to overcome obstacle #1) [20]. Participants completed an implementation intention for each of three obstacles they identified.

2.3.1 Statistical analysis

Apriori power analysis on 52 participants total (26 per group) were calculated to be adequate to detect a group difference (80% power) with an effect size of approximately 0.8 (alpha = 0.05) based on the results of total daily steps reported in a previous 12-week intervention study [17]. The power analysis was performed using G*Power (Kiel University, Germany). All statistical analyses were performed using SPSS^® software (IBM Version 21.0, IBM, Inc., Armonk, NY), with statistical significance set at p < .05 and a tendency of a potentially- meaningful difference at a range of p = .05 –.10. Repeated measure analysis of variance (ANOVA) was performed to compare differences on average steps between groups (PED and PED+II) and at different time points (Baseline and weeks 4, 8 and 12) during the intervention. Confidence intervals for the mean difference between the two groups were also calculated. Bonferroni corrections were performed to control for family-wise errors. Hedges’ g were calculated for effect size of average steps for both groups separately and combined. Effect size plots were created to understand directional differences between interventions.

3 Results

Two participants were excluded from data analysis due to incomplete data resulting in a sample of 54 participants. Of the 54 participants included in analysis, 87% were female (n = 47) with an average age of 47.7±9.0 years. Twenty-three identified their position at the university as civil service, 17 as academic professional, and 14 as faculty. At the start of the intervention, 33 participants rated their overall health status as average, 16 rated their health status as excellent or good, and 5 rated their health status as bad or poor. The average baseline BMI was 31.9±6.0 k/m². See Table 1 for participant demographics.

Table 1
Demographic characteristics of the sample (N = 54)

% Women in sample 87 (n = 47)

Mean age (yrs) 47.70 (SD = 9.019)

Mean BMI (Wk 0) 31.84 (SD = 5.96)

Mean BMI (Wk 8) 31.58 (SD = 5.93)

Employment (%)

Faculty 25.9 (n = 14)

Academic Professional 31.5 (n = 17)

Civil Service 42.6 (n = 23)

Race (%)

Caucasian 90.7 (n = 49)

Black/African American 7.4 (n = 4)

Hispanic/Latino 1.9 (n = 1)

Health status (%)

Poor 1.9 (n = 1)

Bad 7.4 (n = 4)

Average 61.1 (n = 33)

Good 25.9 (n = 14)

Excellent 3.7 (n = 2)

% Women in sample	87 (n = 47)
Mean age (yrs)	47.70 (SD = 9.019)
Mean BMI (Wk 0)	31.84 (SD = 5.96)
Mean BMI (Wk 8)	31.58 (SD = 5.93)
Employment (%)
Faculty	25.9 (n = 14)
Academic Professional	31.5 (n = 17)
Civil Service	42.6 (n = 23)
Race (%)
Caucasian	90.7 (n = 49)
Black/African American	7.4 (n = 4)
Hispanic/Latino	1.9 (n = 1)
Health status (%)
Poor	1.9 (n = 1)
Bad	7.4 (n = 4)
Average	61.1 (n = 33)
Good	25.9 (n = 14)
Excellent	3.7 (n = 2)

Overall significant differences were observed between time points (p < .0001) but not between groups. Post-hoc pairwise comparisons between the different time points revealed difference between baseline and Week 4 (mean difference: 2446.9 steps/ day; p < 0.001), Week 4 and 12 (mean difference: 2956.3 steps/ day; p < 0.001), and Week 8 and 12 (mean difference: 2228.8 steps/ day; p = 0.005). See Tables 2 and 3.

Table 2

Average steps per day for the two intervention groups separately and combined

	Baseline	Week 4	Week 8	Week 12	p value	Grand Mean	Confidence Intervals		Observed Power
							Lower Bound	Upper Bound
Overall	7361.2±2680.2^a	9763.4±3145.6^ab	9060.3±4743.6^c	6825.0±2143.1^bc	<0.0001^*	8226.2	7283.2	9169.2	0.995
PED	7755.1±2892.4	9800.1±2803.2	9291.5±5361.9	7004.4±1951.1	0.613	8462.8	– 1412.8	2359.2	0.10
PED+II	6868.8±2387.2	9717.6±3623.6	8771.3±3990.9	6600.7±6825.0		7989.6	– 2359.2	1412.8

^*Groups with superscripts are significantly different.

Table 3

Time point comparison for average steps per day for both groups combined

(I) Time	(J) Time	Mean Difference (I-J)	% change	Std. Error	p value	95% Confidence Interval for Difference		Bias Corrected Hedges’g
						Lower Bound	Upper Bound
Baseline	Week 4	2446.9^*	31.9	342.8	0.000*	1486.7	3407.2	0.79
Baseline	Week 8	1719.5	22.3	739.6	0.157	– 352.4	3791.4	0.45
Baseline	Week 12	– 509.3	– 9	331.1	0.799	– 1436.9	418.2	– 0.27

The predominantly positive effects of the interventions on average steps improvement at Week 4 were supported by high effect sizes for overall (Hedges g = 0.79; 95% CI=0.40, 1.18) and the two groups individually as well PED (Hedges g = 0.68; 95% CI=0.14, 1.21) and PED+II (Hedges g = 0.88; 95% CI=0.31, 1.45). Moderate effect sizes were also observed at Week 8 (overall (Hedges g = 0.45; 95% CI=0.07, 0.83); PED (Hedges g = 0.42; 95% CI=– 0.11, 0.95) and PED+II (Hedges g = 0.49; 95% CI=– 0.06, 1.04)) but not at Week 12 (Fig. 2). This might be mainly due to lack of adherence the post- 8 week intervention period amongst individuals of both groups. No significant differences were observed between intervention types (groupwise; p = .613). See Fig. 2 for effect sizes.

Fig.2

Effect size plot for average steps per day for the two groups separately and combined.

4 Discussion

4.1 Adoption

Overall, there was an increase in steps from baseline to Week 8 for all participants, although not statistically significant (p = 0.157). Looking more closely at the different time points, there was a statistically significant increase in steps from Baseline to Week 4 (p < 0.001) although this increase was not continued from Week 4 to Week 8. The second phase of the intervention did not result in significant increases in steps like the first phase. Larger increases in step behavior during the first part of a pedometer intervention have been found in other research as well [21]. A reason for this difference may be the format of this intervention. Weeks 0–4 were the Participant Engagement Stage of the intervention. Participants attended weekly meetings in which they interacted with fellow program participants and engaged in behavior modification activities. See Fig. 1 for list of activities participants engaged in each week. This was the most successful phase of increasing steps during the intervention. Weeks 5–9 were the Participant Autonomous Stage of the intervention. Participants had weekly email reminders to meet their weekly step goals but did not attend meetings to hold them accountable. Research has shown that behavior adoption is more successful when intervention engagement is highest [22]. Furthermore, the social interactions of the weekly meetings may have impacted activity levels as well. A program to increase physical activity among employees at small and medium sized enterprises used a peer training program to encourage and promote physical activity [23]. The combination of the high engagement of this phase of the program plus the weekly social interactions may have contributed to the differences in steps across the intervention.

There were no statistical differences between groups but there were differences within the groups. There was a medium effect size for the PED Only group from Baseline to Week 4 compared to the large effect size of the PED+II group during the same time points. Similarly, the PED+II group had a larger effect size from Baseline to Week 8 compared to the PED Only group. These larger effect sizes for the PED+II group indicate a stronger treatment effect on steps during different phases of the intervention. Other studies examining the effects of implementation intentions on behavioral outcomes have found similar group differences [13 , 23]. Specifically, one study examining the effect of implementation intentions on women’s walking behavior found a statistically significant increase in steps in the group that were asked to form specific walking intention plans compared to the group that did not [21]. It concluded that the implementation intentions were an effective strategy for initiating walking within their sample. In the current study, participants wrote implementation intentions addressing a range of potential barriers and solutions such as time management (“If I get up late, then I will walk at lunch”), family responsibilities (“If family obligations prevent me from exercise, then I will incorporate family into exercise”), and work conflicts (“If I feel pushed because of the work I need to accomplish, then I will let something else go, such as reading the newspaper, so that I have time for activity”). This behavioral strategy may explain the strong treatment effects found in the PED+II group compared to the PED Only. This finding may demonstrate that behavioral strategies are pertinent for eliciting changes in physical activity. Future programs to promote physical activity should have participants address implementation intentions dealing with barriers and solutions. Adding a component such as this one to a physical activity promotion program may help participants become more active.

Results of the present study add to our understanding of how implementation intentions effect participants’ behavioral strategies. Often in the literature, when implementation intentions were tested, they were the only behavioral strategy utilized [13 , 28] or combined with the use of self-monitoring through an activity tracking device [21]. The present study tested the utility of implementation intentions in an intervention in which participants were taught multiple behavioral strategies. At the weekly meetings during the Participant Engagement Stage, all participants engaged in behavioral modification activities addressing goal setting and modification, relapse-prevention in addition to identifying motivators and barriers. This multifaceted intervention approach has been successfully utilized in research examining ways to increase physical activity and adherence [25]. The implementation intentions did produce larger step increases in the PED+II group compared to the PED Only group but not statistically significant increases as in other literature [21]. The combination of the implementation intentions with other behavior modification strategiesmay be a reason why the differences between the groups is not as large as reported in other research. When creating interventions, practitioners may consider combining other behavioral strategies with implementation intentions to increase likelihood of behavioral adoption such as self-monitoring through the use of activity trackers.

4.2 Adherence

For both groups, there was a statistically significant decrease in steps from Week 4 to the delayed-post assessment at Week 12 and from the end of the intervention at Week 8 and Week 12. There were no statistical differences between the groups but group differences did emerge. The effect sizes were smaller in the PED+II group compared to the PED Only group when examining Baseline to Week 12 indicating less of a drop off in the PED+II group compared to the PED Only group. In a study examining the use of implementation intentions on adherence to a strength and condition program, Murray, Fraser, and Rodgers [20] also found a decrease in adherence in the experimental group utilizing implementation intentions and the control group that did not. Similar to the present study; adherence was still better in the group that formed implementation intentions compared to the control group. It was concluded that implementation intentions may help maintain behavioral adherence [20].

One factor that may have influenced the adherence results could be the number of participants that engaged in the delayed-post assessment. All participants were asked to complete the delayed-post assessment four weeks after the conclusion of the intervention. However, only 36 participants (64%) completed the delayed-post assessment to evaluate adherence. The results might be different if more participants had completed delayed-post assessments. Future interventions should include delayed-post assessments when evaluating the utility of implementation intentions to promote behavioral adoption and adherence.

4.3 Limitations

A key limitation of the current study was the reliance on a homogenous sample in regards to race/ethnicity and sex. Efforts were made to recruit a more representative sample of the campus employment population. Nevertheless, it is possible that participants that enrolled in the intervention had more favorable views of physical activity and health outcomes than those that chose not to participate. Given the modest sample size, any attempt to generalize these findings to other populations should be avoided. A novel contribution of the present study was testing the utility of implementation intentions when combined with other behavioral strategies. In previous research when implementation intentions were tested, they were the only behavioral strategy utilized [13, 24] or combined with the use of self-monitoring through an activity tracking device [21]. However, in the present study a control group receiving no behavioral strategies may have illuminated the impact of implementation intentions in a variety of conditions within one intervention. Future research should continue to combine multiple behavioral modification strategies for effective and sustained behavioral change.

5 Conclusion

The overarching goal of this investigation was to better understand the impact of implementation intentions on step averages when used with an activity monitor in a worksite setting. To that end, the goal has been accomplished. Implementation intentions have been utilized in various physical activity interventions [26] but have not be used with activity monitors in a worksite setting. The treatment effect is supported in the literature in other interventions promoting various types of behavior modification including smoking cessation [26] and increasing fruit consumption [28]. Specifically, research supports the use of these implementation intentions in worksite wellness interventions suggesting the strategy is effective for this setting. Overall, the results indicate that the intervention produced increases in activity, however; those increases were not maintained indicating that the intervention was effective but adherence is still difficult. The delayed-post assessment is a valuable tool in behavior change research. It is a way to determine the utility of the intervention after the intervention ended. Our results indicated that participants did not maintain the progress made during the intervention four weeks later. It was found that the participants in the PED+II group had slightly more favorable delayed-post assessment results, mitigating the decrease in steps of the PED Only group four weeks after the intervention. Implementation intentions are one behavioral strategy that researchers can use to increase adoption and adherence. Future research should examine other worksite behavior intervention strategies combined with pedometer use to help increase physical activity, goal achievement, and adherence.

Conflict of interest

The authors do not report any conflicts of interest.

Footnotes

Acknowledgments

This work was supported by a grant (grant #512009890) from the Illinois Association for Health, Physical Education, Recreation, and Dance (IAHPERD) Jump Rope for Heart Research Grant.

The results and views of the current study do not constitute endorsements by the American College of Sports Medicine.

References

Bureau of Labor Statistics. [Internet]. American time use survey – 2015 results. 2016 [cited 2018 Jan 12]. Available from: https://www.bls.gov/news.release/pdf/atus.pdf.

Das

, Mailey

, Murray

, et al. From sedentary to active: Shifting the movement paradigm in workplaces. Work. 2016;54(2):481–7.

Bureau of Labor Statistics. [Internet]. Labor force statistics from the current population survey. [cited 2018 Jan 12]. Available from: https://data.bls.gov/timeseries/LNS12300000

Baicker

, Cutler

, Song

. Workplace wellness programs can generate savings. Health Aff (Millwood). 2010;29:304–11.

Pinkstaff

, Arean

, Myers

, et al. The Affordable Care Act: New opportunities for cardiac rehabilitation in the workplace? J Occup Environ Med. 2014;56(8):809–13.

Center for Disease Control and Prevention [Internet]. Atlanta (GA): Facts about physical activity; 2014 [cited 2018 Jan 12]. Available from: https://www.cdc.gov/physicalactivity/data/index.html.

American Heart Association. [Internet]. The price of inactivity. 2015 [cited 2018 Jan 12]. Available from: http://www.heart.org/HEARTORG/HealthyLiving/PhysicalActivity/FitnessBasics/The-Price-of-Inactivity_UCM_307974_Article.jsp.

Kohl

, Craig

, Lambert

, et al. The pandemic of physical inactivity: Global action for public health. Lancet. 2012;380(9838):21–7.

Lee

, Shiroma

, Lobelo

, et al. Impact of physical inactivity on the world’s major non-communicable diseases. Lancet. 2012;380(9838):219–29.

10.

Tremblay

, Colley

, Saunders

, et al. Physiological and health implications of a sedentary lifestyle. Appl Physiol Nutr Metab. 2010;35:725–40.

11.

Dishman

. The problem of exercise adherence: Fighting sloth in nations with market economics. Quest. 2001;53:279–94.

12.

Sheeran

, Webb

, Gollwitzer

. Implementation intentions: Strategic automatization of goal striving. In: Ridder

, Wit

, editors. Self-regulation in health behavior. West Sussex, England: John Wiley and Sons Ltd; 2006, pp. 121–45.

13.

Murray

, Rodgers

, Fraser

. Examining implementation intentions in an exercise intervention: The effects on adherence and self-efficacy in a naturalistic setting. J Appl Soc Psychol. 2009;39(10):2303–20.

14.

Kang

, Marshall

, Barreira

, et al. Effect of pedometer-based physical activity interventions: A meta-analysis. Res Q Exerc Sport. 2009;80(3):648–55.

15.

Bravata

, Smith-Spangler

. Sundaram

, et al. Using pedometers to increase physical activity and improve health: A systematic review. JAMA. 2007;298(19):2296–304.

16.

Tudor-Locke

, Lutes

. Why do pedometers work? A reflection upon the factors related to successfully increasing physical activity. Sports Med. 2009;39(12):981–93.

17.

Croteau

. Using pedometers to increase the non-workday steps of hospital nursing and support staff: A pilot study. Workplace Health Saf. 2017;65(10):452–6.

18.

Center for Disease Control and Prevention [Internet]. Atlanta (GA): National diabetes prevention program. 2016. [cited 2018 Jan 12]. Available from: https://www.cdc.gov/diabetes/prevention/index.html.

19.

Deci

, Ryan

. Intrinsic motivation and self-determination in human behavior. New York, NY: Plenum Publishing Co; 1985.

20.

Stadler

, Oettingen

, Gollwitzer

. Physical activity in women: Effects of a self-regulation intervention. Am J Prev Med. 2009;36(1):29–34.

21.

Arbour

, Martin Ginis

. A randomised controlled trial of the effects of implementation intentions on women’s walking behaviour. Psychol Health. 2009;24(1):49–65.

22.

Kolt

, Rosenkranz

, Vandelanotte

, et al. Using Web 2.0 applications to promote health-related physical activity: Findings from the WALK 2.0 randomised controlled trial. Br J Sports Med. 2017;0:1–8.

23.

Edmunds

, Duncan

, Clow

. The effects of a physical activity intervention on employees in small and medium enterprises: A mixed methods study. Work. 2013;46(4):39–49.

24.

Luszczynska

, Sobczyk

, Abraham

. Planning to Lose Weight: Randomized controlled trial of an implementation intention prompt to enhance weight reduction among overweight and obese women. Health Psychol. 2007;26(4):507–12.

25.

Rejeski

, Brawley

, Ambrosius

, et al. Older adults with chronic disease: Benefits of group-mediated counseling in the promotion of physically active lifestyles. Health Psychol. 2003;22(4):414–23.

26.

Belanger-Gravel

, Godin

, Amireault

. A meta-analytic review of the effect of implementation intentions on physical activity. Health Psychol Rev. 2013;7(1):23–4.

27.

Armitage

. Evidence that Implementation intentions can overcome the effects of smoking habits. Health Psychol. 2016;35(9):935–43.

28.

Armitage

. Field experiment of a very brief worksite intervention to improve nutrition among health care workers. J Behav Med. 2015;38:599–608.