Social support for physical activity: Comparison of family,friends,and coworkers

Abstract

BACKGROUND: Few studies have examined the associations among family, friend, and coworker social support for physical activity. It is important to know the sources of social support that facilitate and promote physical activity among fulltime working adults.

OBJECTIVE: We analyzed the associations among family, friend, and coworker social support for physical activity and moderate-to-vigorous physical activity among participants in a worksite study.

METHODS: This study was a cross-sectional analysis of baseline data from 144 participants from four worksites in a large, southwestern city in the United States. The intervention for the worksite study was Booster Breaks (a physical activity routine of 15 minutes) practiced daily to break-up prolonged sitting time. Descriptive statistics and multiple linear regressions were conducted using SPSS version 20.

RESULTS: Age was inversely associated (p = 0.001), and social support from friends (p = 0.04) and coworkers (p = 0.003) were positively associated with physical activity in the unadjusted model. After controlling for all the covariates (age, sex, marital status, BMI, education, and income) in the model, only coworker social support was positively (p = 0.027) associated with physical activity among participants in the workplace study.

CONCLUSIONS: Coworker social support is an important correlate of physical activity and should be incorporated in workplace health promotion programs.

Keywords

Workplace health promotion social support physical activity

1 Introduction

Social support for physical activity refers to any form of encouragement from a person’s social network that is positively associated with physical activity [1 –3]. Social support from family and friends has been positively associated with physical activity [4]. Prior studies have found that family social support improved physical activity [5, 6]. Also, having a friend or companion for physical activity was positively associated with neighborhood walking, and recreational and leisure time physical activity [7]. The specific mechanism by which social support improves physical activity is not clear. However, some studies have indicated that social support increases the motivation and self-efficacy (i.e., confidence in ability) to participate in moderate-to-vigorous physical activity despite barriers and thus, these psychosocial factors are likely to improve physical activity participation [2, 3].

Previous studies have found that increased sedentary (sitting) time is associated with adverse health outcomes such as overweight, obesity, and chronic diseases like metabolic syndrome, cardiovascular diseases, type 2 diabetes mellitus, coronary artery diseases, and cancer including increased morbidity and mortality [8 –11]. Some studies have emphasized the beneficial effects of breaking up sedentary time and adherence to moderate-to-vigorous intensity physical activity. Breaking up sedentary time has been associated with lower waist circumference, improvement in serum triglyceride and lower blood glucose levels [5 , 12–14]. Nevertheless, the majority of physical activity studies have failed to address the psychosocial correlates of physical activity and fewer studies have examined the importance of coworker social support in workplace studies [2 , 15].

Studies examining the association between employee social support and physical activity found that individuals with greater mean scores for workplace social support had greater mean physical activity scores [16, 17]. However, to our knowledge no study has compared the extent and type of social support for physical activity from coworkers to that of friends and family. The extent and type of social support can be quantified by combining the degree of frequency (e.g., never to very often) with different types of social support (encouragement, participation, and willingness to help) and then analyze the relationship with physical activity levels. Because of frequent contact, communication, and the potential substantial influence of coworkers, we hypothesized that employees with greater coworker social support for physical activity compared to family and friends will have greater levels of physical activity. The research question was: Who are important sources of social support for physical activity when comparing family, friends, and coworkers?

Social support is an important construct for physical activity interventions. Our study will provide new data on the influence of coworker social support for physical activity after adjusting for other forms of social support (i.e., family and friends). This paper represents a dramatic shift from prior studies that examined the impact of various forms of social support in isolation. Our approach is to assess the relative importance of each source of social support. From a public health perspective, knowing the relevance of each source of social support can inform intervention development.

2 Methods

2.1 Booster Breaks

Booster Breaks are defined as “organized, routine work breaks intended to improve physical and psychological health, enhance job satisfaction, and sustain or increase work productivity”[18]^(p . 462). The Booster Break is considered an alternative to typical work break practices such as smoking cigarettes, drinking coffee, consuming high calorie drinks or snacks from vending machines, and surfing the internet [18, 19]. Booster Breaks are 15-minute sessions designed to break up extended periods of sedentary time [20]. Booster Break facilitators are coworkers trained and certified to lead the workplace group physical activity sessions [20]. The facilitators guide and direct the physical activity routine per protocol as described in an earlier publication [20].

2.1.1 Study design, sample, and setting

This cross-sectional analysis was from baseline data of Booster Break participants who completed surveys (2009-2010) administered at their workplaces. Employees (men and women) whose job required sitting for at least 5 hours per day from four worksites (department of education, hospital, law firm, and court reporting/video agency) were included in this study.

2.1.2 Procedure

An analysis of participants’ baseline responses to the Booster Break Survey was performed for demographic, physical activity, and social support variables. Participants’ responses to the Booster Break Survey for baseline social support and physical activity were entered in the data editor of SPSS version 20. Constructs of this survey are described below.

2.2 Physical activity measure

Physical activity (dependent variable) was quantitative and measured as duration of physical activity in minutes per day for the previous seven days. Physical activity questions in the Booster Break Survey were from the long version of the International Physical Activity Questionnaire (IPAQ) which included four domains of physical activity [21]. Physical activity data were a continuous variable obtained by summing up the duration of physical activity from items listed in question E (Job related physical activity) + question F (Transportation physical activity) + question G (Housework related physical activity) + question H (Recreation and Leisure time physical activity).

The Booster Break Survey physical activity questions were self-administered by the participants. The IPAQ has been used for physical activity measurement in many published studies [21, 22] and is a reliable assessment [21 –24]. For the IPAQ, construct and concurrent validity have been reported [21, 25]. For example, concurrent validity was demonstrated for moderate-to-vigorous intensity physical activity and accelerometer-determined physical activity with significant Spearman’s correlations rho = 0.36 [21].

2.3 Social support measure

Social support for physical activity was measured (question item Q of the Booster Break Survey) using an adapted version of Social Support for Exercise Survey [26, 27]. Support from family, friends, and coworkers were assessed separately, by asking the respondent “During the past three months my family, friends, and coworkers: 1) Did physical activity with me; 2) Offered to do physical activity with me; 3) Gave me encouragement to do physical activity.” Responses were elicited separately from family, friends, and coworkers using a five-point scale (0 = Never; 4 = Very often). Responses from the three questions were then summed separately for family, friends, and coworkers social support.

2.3.1 Data analysis

Descriptive statistics (mean, median, range, standard deviation) for the study variables were computed (Table 2). Spearman’s correlations were calculated among the study variables to account for the non-normal distributions (Table 3).

Data analyses were conducted in two steps. First, simple linear regressions were performed separately between the independent variables (family, friends, and coworkers social support) and self-reported physical activity (dependent variable) entered one at a time in the regression model to evaluate their linear relationship (Table 4). Later multiple linear regressions were performed between social support (family, friends, and coworkers) and self-reported physical activity after controlling for covariates (age, sex, marital status, BMI, education, and income) in the model which were potential confounders (Table 5). The change in variance (R²) in physical activity (outcome variable) along with significance levels (p < 0.05) were examined.

3 Results

3.1 Demographic data

The mean age (SD) for the participants was 44 (±13) years. The socio-demographic characteristics indicated that 117 (81%) of the participants were females, 95 (66%) had college level education, and 91 (63%) were married or living with a partner. Approximately, 40 (28%) participants had an annual household income that ranged between 50,000–79,000 dollars (Table 1).

3.2 Descriptive data

Descriptive statistics for the study variables indicated that the mean (SD) for BMI were 29.7 (±7) and for physical activity were 2223.55 (±2275.62) minutes per week (Table 2). Physical activity (dependent variable) was non-normal in distribution. For physical activity, the values were 900 (25%), 1275 (50%), and 2540 (75%) minutes per week. The values for family social support were 4.73 (±3.43); friend social support were 3.46 (±3.33) and for coworker social support were 2.79 (±3.09). The family, friend, and coworker social support scores ranged from 0–12. Social support (independent variable) was also non-normal in distribution. The (25th, 50th, 75th percentiles) for family social support scores were 2, 4, and 7, friend social support were 0, 3, and 6, and coworker social support were 0, 2, and 4.

Physical activity and social support data were non-normal in distribution (Shapiro Wilk test p < 0.05) and were positively skewed on the histogram. A log transformation of the outcome variable (physical activity) was performed to normalize the data for regression analysis.

Age was significantly and inversely correlated with physical activity, friend, and coworker social support and had low-to-moderate Spearman’s correlation coefficients (rho = –0.26, –0.37, –0.27, respectively with each variable; p = 0.01) (Table 2). Family social support had significant positive correlations with friend and coworker social support and low-to-moderate Spearman’s correlation coefficients (rho = 0.32, 0.28 respectively; p = 0.01). Friend social support had a significant inverse correlation with age (rho = –0.37, p = 0.01) and a significant positive correlation with physical activity (rho = 0.19, p = 0.05). Friend social support had significant positive correlations with family and coworker social support (rho = 0.32, 0.66 respectively; p = 0.01). Coworker social support had a significant and inverse correlation with age (rho = –0.27, p = 0.01). Coworker social support had a significant and positive correlation with physical activity (rho = 0.31, p = 0.01) and significant and positive correlations with friend and family social support (rho = 0.66, 0.28 respectively; p = 0.01).

A simple linear regression performed between physical activity and the study variables indicated that age was significantly and inversely associated (β= –0.278; p = 0.001) with physical activity, whereas social support from friends (β= 0.172; p = 0.04) and coworkers (β= 0.248; p = 0.003) were significantly and positively associated with physical activity (Table 4). Based on the patterns of inter-correlations among the study variables and linear regression analysis, a multiple regression analysis was performed. The observed data did not significantly deviate from the assumptions of the regression model.

After controlling for the covariates (age, sex, marital status, education, income, and body mass index), the predictor variable (coworker social support) was significantly associated with physical activity (β= 0.269; p = 0.027) (Table 4). Family and friend social support were not significantly associated with physical activity in the model. Although friend social support initially along with coworker social support had a significant association with physical activity in the linear regression analysis (Table 4), however, after controlling for the covariates in the model, only coworker social support was significantly associated with physical activity (Table 5). The R² value of 0.166 indicated that 17% of the variance in physical activity behavior was attributable to the predictor variables: family, friend, and coworker social support, in the model after controlling for age, sex, marital status, education, income, and body mass index. Only coworker social support was statistically significant (p = 0.027).

In the first step, we included control variables (age, sex, marital status, education, income, and body mass index) in the regression model (Table 5). Then in the next step the predictor variables (family, friend, and coworker social support) were added. There was a 5% change in R² observed from step 1 to step 2 which was statistically significant (p = 0.008).

4 Discussion

The results indicated that friend social support along with coworker social support had a significant association with physical activity; however, after controlling for the covariates in the adjusted model, coworker social support was the only source of support that was significantly associated with physical activity. These data suggest that coworker social support is an important correlate of physical activity that transcends the work environment and contributes to overall physical activity.

This finding is consistent with other workplace physical activity studies that found a positive association between worksite social support and physical activity [16, 17]. One worksite study [17] conducted among women employed in rural manufacturing worksites in North Carolina found that workplace social support among women was associated with not only physical activity but also other health promotion behaviors (e.g., fruit and vegetable intake, and cervical cancer screening). The study emphasized that women who had larger social networks and higher levels of coworker interactions engaged in more health promotion behaviors [17]. Another workplace study [28] found that coworker physical activity participation provided social support and was significantly associated with objectively measured pedometer based physical activity. Our study is unique because we examined coworker social support in the context of other forms of social support to determine relative contributions of social support. Prior studies did not report comparisons by sources of social support [16 , 28].

One possible explanation for why coworker social support was a greater predictor of physical activity is that employees spend a substantial amount of time (8 or more hours per day; 40 or more hours per week) with coworkers and are likely to respond favorably to social support provided by coworkers. Another study [29] emphasized encouragement from peer, physical activity champions, and supportive social networks in the workplace and found increased employee physical activity. Further studies are needed to identify the forms of social support that are most relevant for physical activity during the work day and outside of the workplace. The four functional domains of social support are appraisal (perceived availability of someone to talk to about one’s problems), belonging (emotional availability of people one can perform the activity with), tangible (instrumental support or material aid), and self-esteem (validation or comparing self to others) [3, 4]. Future research can investigate social support from a theoretical perspective to better understand the nature of the effects.

4.1 Limitations

The Booster Break Survey did not include questions about contact time for each source of social support (i.e., duration of time spent with friends, family, and coworkers). We recommend that future research include time spent with friends, family, and coworkers as well as the context of the contact time. The addition of this variable in future research can provide interesting analyses to identify the mechanisms of influence.

Employees volunteering to participate in a workplace physical activity study may have greater interest in physical activity and thus, the possibility of selection bias. Furthermore, the participants in the Booster Break workplace study had desk jobs and most of the participants were women (81%). Therefore, the generalizability of these results to other occupations and men is unknown and merits further study. Also, self-report measures may introduce recall or respondent bias such as over-reporting physical activity [30]. Finally, data on the predictor and outcome variables were collected at one moment in time. Because this study was cross-sectional, we are unable to make any conclusions regarding causality.

4.2 Strengths

In spite of the limitations, the study has several strengths. The study sample was composed of a racially and ethnically diverse population. This profile is important because of the increasing racial and ethnic diversity of the US population and workforce [31, 32]. Importantly, this research compared social support among family, friends, and coworkers and examined these associations with physical activity. No known study has compared different sources of social support for physical activity.

Our findings enable us to better understand the correlates of physical activity. Prior studies have emphasized the role of family and friend social support for physical activity among various populations [1 , 33]. In contrast, our findings emphasize the significance of coworker social support for physical activity after adjusting for other sources of social support and confounders. These data can inform workplace program designers about the importance of coworker social support for physical activity.

5 Conclusions

Coworker social support is a significant correlate of physical activity. Physical activity programs should consider the importance of coworker social support in designing workplace programs. Therefore, this study is a call to action to design workplace physical activity programs and advocate for policies directed towards participation in workplace programs by capitalizing on coworker social support.

Future research can identify how best to mobilize coworker social support and what types of coworker social support are most effective. In a single-arm repeated measures design, workplace activators (i.e., natural helpers) encouraged their peers to participate in physical activity opportunities by promoting companionship and social support via social networks in the workplace [34]. We recommend that randomized controlled trials be conducted to operationalize coworker social support for physical activity in different ways and evaluate the outcomes to advance the field of workplace health promotion.

Conflict of interest

None to report.

Footnotes

Acknowledgments

The research was supported by grant number, R03 NR010291, from the National Institutes of Health (NIH) awarded to Wendell C. Taylor, PhD, MPH. The contents of the publication are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. This paper originated from the dissertation research of Dr. Sonali Sarkar. Her dissertation committee members were Drs. Wendell C. Taylor (chair) and Dejian Lai.

References

King

, Tergerson

, Wilson

. Effect of social support on adolescents’ perceptions of and engagement in physical activity. J Phys Act Health 2008;5:374–84.

Anderson

, Wojcik

, Winett

, Williams

. Social-cognitive determinants of physical activity: The influence of social support, self-efficacy, outcome expectations, and self-regulation among participants in a church-based health promotion study. Health Psychol 2006;25(4):510–20.

Cohen

, Mermelstein

, Kamarck

, Hoberman

. Measuring the functional components of social support. In: Sarason

, Sarason

, editors. Social support: Theory, research and application. The Hague, Holland: Martinus Nijhoff; 1985: pp. 73–94.

Duncan

, Duncan

, Strycker

. Sources and types of social support in youth physical activity. Health Psychol 2005;24(1):3–10.

Khan

, Stephens

MAP

, Franks

, Rook

, Salem

. Influences of spousal support and control on diabetes management through physical activity. Health Psychol 2013;32(7):739–47.

Shelton

, McNeill

, Puleo

, Wolin

, Emmons

, Bennett

. The association between social factors and physical activity among low income adults living in public housing. Am J Public Health 2011;101(11):2102–10.

Wendel-Vos

, Droomers

, Kremers

, Brug

, Van Lenthe

. Potential environmental determinants of physical activity in adults: A systematic review. Obes Rev 2007;8:425–40.

Bankoski

, Harris

, McClain

, Brychta

, Caserotti

, Chen

, et al. Sedentary activity associated with metabolic syndrome independent of physical activity. Diabetes Care 2011;34(2):497–503.

Healy

, Matthews

, Dunstan

, Winkler

E A

, Owen

. Sedentary time and cardio-metabolic biomarkers in US adults: NHANES 2003–06. Eur Heart J 2011;32(5):590–7.

10.

Owen

, Sparling

, Healy

, Dunstan

, Matthews

. Sedentary behavior: Emerging evidence for a new health risk. Mayo Clin Proc 2010;85(12):1138–41.

11.

Katzmarzyk

, Church

, Craig

, Bouchard

. Sitting time and mortality from all causes, cardiovascular disease, and cancer. Med Sci Sports Exerc 2009;41(5):998–1005.

12.

Cooper

, Sebire

, Montgomery

, Peters

, Sharp

, Jackson

, et al. Sedentary time, breaks in sedentary time and metabolic variables in people with newly diagnosed type 2 diabetes. Diabetologia 2012;55(3):589–99.

13.

Dunstan

, Kingwell

, Larsen

, Healy

, Cerin

, Hamilton

, et al. Breaking up prolonged sitting reduces postprandial glucose and insulin responses. Diabetes Care 2012;35(5):976–83.

14.

Healy

, Dunstan

, Salmon

, Cerin

, Shaw

, Zimmet

, et al. Breaks in sedentary time beneficial associations with metabolic risk. Diabetes Care 2008;31(4):661–6.

15.

Carlson

, Sallis

, Conway

, Saelens

, Frank

, Kerr

, et al. Interactions between psychosocial and built environment factors in explaining older adults’ physical activity. Prev Med 2012;54(1):68–73.

16.

Tamers

, Beresford

SAA

, Cheadle

, Zheng

, Bishop

, Thompson

. The association between worksite social support, diet, physical activity and body mass index. Prev Med 2011;53(1-2):53–6.

17.

Kelsey

, Campbell

, Tessaro

, Benedict

, Belton

, Fernandez

, et al. Social support and health behaviors among blue-collar women workers. Am J Health Behav 2000;24(6):434–43.

18.

Taylor

. Transforming work breaks to promote health. Am J Prev Med 2005;29(5):461–5.

19.

Taylor

. Booster Breaks: An easy-to-implement workplace policy designed to improve employee health, increase productivity, and lower health care costs. J Workplace Behav Health 2011;26(1):70–84.

20.

Taylor

, Shegog

, Chen

, Rempel

, Pappas-Baun

, Bush

, et al. The Booster Break Program: Description and feasibility test of a worksite physical activity daily practice. Work 2010;37:433–43.

21.

Hagstromer

, Oja

, Sjostrom

. The International Physical Activity Questionnaire (IPAQ): A study of concurrent and construct validity. Public Health Nutr 2005;9(6):755–62.

22.

Fogelholm

, Malmberg

, Suni

, Santtila

, Kyrolainen

, Mantysaari

, et al. International physical activity questionnaire: Validity against fitness. Med Sci Sports Exerc 2006;38(4):753–60.

23.

Lee

, Macfarlane

, Lam

, Stewart

. Validity of the international physical activity questionnaire short form (IPAQ-SF): A systematic review. Int J Behav Nutr Phys Act 2011;8:115 Available from: http://www.ijbnpa.org/content/doi./10.1186/1479-5868-8-115

24.

Craig

, Marshall

, Sjostrom

, Bauman

, Booth

, Ainsworth

, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003;195:3508–1381.

25.

Pereira

, FitzerGerald

, Gregg

, Joswiak

, Ryan

, Suminski

, et al., eds. A collection of Physical Activity Questionnaires for health-related research [special issue]. Med Sci Sports and Exerc 1997;29:S1–205.

26.

Sallis

, Grossman

, Pinski

, Patterson

, Nader

. The development of scales to measure social support for diet and exercise behaviors. Prev Med 1987;16:825–36.

27.

Sallis

, Buono

, Roby

, Micale

, Nelson

. Seven-day recall and other physical activity self-reports in children and adolescents. Med Sci Sports Exerc 1993;25(1):99–108.

28.

Lauzon

, Chan

, Myers

, Tudor-Locke

. Participant exercises in a workplace pedometer based physical activity program. J Phys Act Health 2008;5(5):675–87.

29.

Edmunds

, Clow

. The role of peer physical activity champions in the workplace: A qualitative study. Perspect Public Health 2015; Sept 16 pii 1757913915600741. [Epub ahead of print].

30.

Rzewnicki

, Auweele

, Bourdeaudhuij

. Addressing overreporting on the International Physical Activity Questionnaire (IPAQ) telephone survey with a population sample. Public Health Nutr 2003;6(03):299–305.

31.

Perez

, Hirschman

. The changing racial and ethnic composition of the US population: Emerging American identities. Popul Dev Rev 2009;35:1–51.

32.

Alba

, Yrizar Barbosa

. Room at the top? Minority mobility and the transition to demographic diversity in the USA. Ethn Racial Stud 2015;1–22.

33.

King

. Interventions to promote physical activity by older adults. J Gerontol A Biol Sci Med Sci 2001;56(2):36–46.

34.

Edmunds

, Stephenson

, Clow

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