- Email: [email protected]
Use of Text Messaging Services to Promote Health Behaviors in Children
Research Brief Use of Text Messaging Services to Promote Health Behaviors in Children nia Gonc¸alves, PhD2; Daniel B. Fassnacht, PhD1,2; Kathina Ali, BSc3; C atia Silva, MSc2; So Paulo P. P. Machado, PhD2 ABSTRACT Objective: To examine adherence to, satisfaction with, and preliminary efficacy of mobile phone short message service (SMS) to promote health behaviors in school-aged children. Methods: A total of 49 children (aged 8–10 years) were randomized by school classes into a monitoring vs nomonitoring group. All children participated in 2 educational group sessions that focused on health behaviors: the advantages of increasing fruit and vegetable consumption and physical activity, and decreasing screen time. The monitoring group also reported daily behavior using SMS and received supportive feedback for 8 weeks. Results: Children submitted 61% of the required SMS, which indicated good adherence to the intervention. A number of children (95%) reported being satisfied with the program. Analyses of covariance indicated increase in fruit and vegetable consumption (c2 [2] ¼ 7.27; P < .05) and a decrease in screen time (c2 [2] ¼ 6.79; P < .05). Conclusions and Implications: The current SMS intervention was a useful tool to monitor and promote health behaviors in children. Key Words: Health behavior, eating behavior, exercise, text messaging (J Nutr Educ Behav. 2015;47:75-80.) Accepted August 11, 2014. Published online October 2, 2014.
INTRODUCTION The rise in childhood obesity has been drastic over the past decades and is one of the most serious public health challenges of the 21st century.1 According to the International Association for the Study of Obesity, over 200 million school-aged children are overweight, which makes this generation the first predicted to have a shorter lifespan than their parents.2 Childhood obesity is mainly associated with unhealthy diets and low physical activity. Health promotion and intervention programs for childhood obesity need to support and facilitate an increase in physical activity and healthier diets. The use of new information and communication technologies has
shown promising results in promoting health behaviors,3 preventing childhood obesity,4 and maintaining weight loss in overweight children.5 For example, the short message service (SMS) has proven to be useful in promoting health behaviors in children.6 In fact, SMS has been acceptable for providing support, effecting behavior change, and/or maintaining treatment gains in diabetes,7 asthma,8 smoking cessation,9 depressive symptoms,10 bulimia nervosa,11 and childhood obesity.12 Children prefer novel devices, and technology-enhanced systems showed positive effects on participants' adherence rates.6 Furthermore, SMS has several advantages such as accessibility at any time, fast and instant information, guidance and
1
Department of Psychology, James Cook University, Singapore Psychotherapy and Psychopathology Research Unit, CIPsi, School of Psychology, University of Minho, Braga, Portugal 3 Centre for Mental Health Research, Australian National University, Canberra School of Psychology, Campus de Gualtar, Braga, Portugal Address for correspondence: Paulo P. P. Machado, PhD, School of Psychology, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal; Phone: (þ351) 253-604220; Fax: (þ351) 253-604224; E-mail: [email protected] Ó2015 SOCIETY FOR NUTRITION EDUCATION AND BEHAVIOR http://dx.doi.org/10.1016/j.jneb.2014.08.006 2
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015
advice without major effort (after messages are created), and low costs.13 The use of Internet and mobile phones has increased in the past 2 decades; according to Cardoso et al,14 84.2% of children aged 9–12 years in Portugal own a mobile phone and 88% use it to type messages. Bauer et al15 developed an Internetbased computer program used in previous research to provide support and promote behavior change via SMS. Methods from social cognitive theory and behavioral models were considered and used in developing the program. Self-monitoring and immediate feedback based on specific goals are important elements in behavioral theory. The idea that health behavior will change with specific goals when positive reinforcement is provided is based on social cognitive theory.16 Shapiro et al6 used an adopted version of the program developed by Bauer et al15 and showed that text messaging might be a useful tool for self-monitoring sugar-sweetened beverages, physical activity, and screen time. Children monitored target behaviors via SMS with feedback or via paper diaries, or participated in a no-monitoring condition. In the current study, the researchers
75
76 Fassnacht et al adapted this program to promote health behaviors (eg, fruit and vegetable consumption, physical activity, screen time) in children. Compared with the study by Shapiro et al, children were randomized by classes into SMS monitoring vs no monitoring.6 The intervention consisted of an interaction between participants and provider via SMS. Participants sent text messages at regular intervals and received supportive feedback messages based on their entries. The current pilot study aimed to explore participants' adherence to and satisfaction with the SMS-based monitoring and feedback system. A secondary aim of the pilot study was to explore the preliminary efficacy of the program to promote health behaviors. To test this aim, researchers hypothesized that the intervention group would report increased fruit and vegetable consumption and physical activity and decreased screen time.
METHODS Children from the fourth grade were recruited from an elementary school in Braga, Portugal. By tossing a coin, the children of 2 school classes were assigned to either a monitoring (intervention: n ¼ 22) or control (n ¼ 27) condition. Children (n ¼ 49) between the ages of 8 and 10 years (mean, 9.6 years; SD, 0.4 years) were included in the study regardless of weight or ethnicity. The intervention with the SMS-based monitoring and feedback system took place over 8 weeks.
Instruments Participants in both groups completed a self-report health behavior questionnaire at the beginning of the first educational group session (baseline), after 8 weeks of monitoring (postintervention), and at 4 weeks' followup in class. The researchers measured usual intake of fruits and vegetables using questions of a paper-based food frequency questionnaire based on the Health Behaviour in SchoolAged Children Study.17 Children were asked, ‘‘How many times a week do you usually eat/drink .’’ followed by a list of food and beverage items including fruits and vegetables. Options were never, less than once a
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 week, once a week, 2–4 d/wk, 5–6 d/wk, once every day, and several times every day. The food frequency questionnaire showed good reliability and validity.18 Daily physical activity and screen time were measured using 2 questions from the Family Eating and Activity Habits Questionnaire19: ‘‘How many hours per day on average did you participate in activities such as fast walking, swimming, ball games?’’; ‘‘How many minutes did you spend in front of the screen (eg, television, computer, video games)?’’ The Family Eating and Activity Habits Questionnaire showed good reliability and validity.19 These questions were used to explore preliminary efficacy. All questionnaires were translated from English into Portuguese and back-translated by the authors. Items were carefully checked by native speakers and changed or edited when necessary. Height and weight were assessed (baseline) without shoes using a digital scale and a stadiometer. Body mass index (kg/m2) for age was calculated for children according to the International Obesity Task Force using SD scores (z-scores) as recommended by Cole.2,20 Children in the intervention group received a pedometer (Silva Pedometer Plus, Silva AB, Bromma, Sweden, 2007) to monitor their steps per day accurately. In addition, children from the intervention group completed a satisfaction questionnaire with 8 questions (eg, ‘‘How much fun did you have doing this program?’’; response options were No fun at all, No fun, Neither . nor, Fun, and Lots of fun) after the intervention to explore children's overall satisfaction with specific components of the program (eg, appropriateness of feedback messages).
Procedure Parents and children received information sheets with necessary details on the planned program (ie, use of mobile phones and pedometers for the intervention group), including informed consent. If parents agreed to their child's participation, they were invited to participate in 1 educational session for parents, were informed about the importance of health behaviors, and were given
detailed information about the program. All children participated in 2 60-minute educational sessions presented in a group format and facilitated by 2 trained psychologists. The only material that differed between the intervention and control groups was presented in session 2, which included detailed information about and training with the SMS program. Session 1 focused on increasing physical activity, decreasing screen time, and the risks of sedentary behavior. Session 2 focused on a healthy diet in general and the importance of the consumption of fruits and vegetables specifically. Children from the intervention group were asked to monitor their fruit and vegetable consumption, physical activity, and screen time daily. Regarding fruit and vegetable consumption, specific serving sizes were discussed and children were asked to report the exact amount of daily fruit and vegetable intake. Children were instructed to report data in a standard format via SMS. Specific questions to obtain this information included ‘‘How many fruits and vegetables did you eat today?’’ ‘‘How many steps did you achieve with the pedometer today?’’ and ‘‘How many minutes did you spend in front of the screen today?’’ To create normative values for the 3 behaviors, the authors used previous research to define 5 portions of fruits and vegetables, healthy amounts of physical activity at 10,000 steps/d, and < 60 min/d screen time. Recommendations for physical activity and sedentary behavior were based on Tudor-Locke et al21,22 and Shapiro et al,6 because national recommendations do not exist specifically for Portugal. Recommendations for fruit and vegetable consumption were based on the World Health Organization, which considers an ‘‘adequate quantity’’ as at least 400–500 g/d, which is equivalent to 5 servings of 80 g of fruits and/or vegetables.23 Children were asked to use their parent's mobile phones, because parents were to support and help the children throughout the program. Both children and parents became acquainted with the appropriate use of the pedometer and how to transfer the data to the daily SMS to submit the SMS together. All costs for mobile
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 phone usage regarding the program were reimbursed; however, no incentives were provided. Messages from participants were sent to a modem connected to a secure Web server. Incoming and outgoing messages were recorded and the software program automatically suggested tailored feedback messages. For the current study, fruit and vegetable consumption replaced consumption of sugar-sweetened beverages. A pool of 900 existing messages from previous research with the SMS program was translated from English into Portuguese and new ones were created to address fruit and vegetable consumption.6 Algorithms for the feedback messages were based on how many of the 3 goals were met and improvement or deterioration from the previous day (eg, ‘‘Great, you met your goal for physical activity and screen time! What happened to fruits and vegetables?’’). The prompted answers were transformed dichotomously by the software program according to whether the behavioral goal was achieved and how this outcome related to the previous status report of the child. Details about the algorithm used and the development of the program were described elsewhere.6,15 The feedback messages aimed to motivate and encourage children to reach each behavioral goal and to support and reinforce positive development based on improved or deteriorated behavior. Giving technological constraints, feedback messages had a limit of 160 characters. The Ethics Committee of the School of Psychology approved the study protocol and all parents/legal guardians signed informed consent after child assent.
time as the within-subjects factor and condition as the betweensubjects factor to compare the amount of consumed fruits and vegetables, physical activity, and screen time between conditions (intervention and control group) at pre- and postintervention and follow-up (time). Gender and baseline levels of the 3 target behaviors were included as covariates. Main effects of time and interaction effects of time group were found. In addition, interaction effects of the covariates (gender, baseline levels of the 3 behaviors) and time were observed. Assumptions of sphericity were tested using the Mauchly sphericity test. For fruit and vegetable consumption (c2 [2] ¼ 7.27; P < .05) and screen time (c2 [2] ¼ 6.79; P < .05), Mauchly tests indicated that the assumption of sphericity had been violated; therefore, degrees of freedom were corrected using Huynh–Feldt estimates of sphericity (ε ¼ .96). Assumptions of normality were tested using the Shapiro–Wilks test. Because of the small sample size, it was not surprising that the assumption for physical activity and screen time was violated for both the control and intervention groups. To test the influence of the normality violation, the data of these 2 variables were both z- and rank-transformed and subsequent ANCOVAs were performed on the transformed data. The 2 procedures gave nearly identical results; hence, the assumptions underlying the ANCOVA are likely to be reasonable.24 In the following only the parametric results of the ANCOVAs are reported. All analyses were performed using SPSS (version 20.0, SPSS Inc, Chicago, IL, 2011).
Data Analysis
RESULTS Participants
The authors measured adherence to the monitoring system by dividing the total number of required monitoring assessments (ie, 56 days) by the number of days participants actually sent their monitored behavior via SMS. To explore the preliminary efficacy of the monitoring and feedback system, the researchers conduced repeated-measures ANCOVA with
Table 1 lists demographic variables and baseline behaviors (fruit and vegetable consumption, physical activity, and screen time) of the sample. A total of 51 children were eligible and 49 participated in the study (2 did not provide signed parental informed consent). The intervention group consisted of significantly more males (n ¼ 14) compared with the control group (n ¼ 9). Children were aged 8
Fassnacht et al 77 to 10 years (mean, 9.6; SD, 0.4). Using body mass index cutoffs provided by the International Obesity Task Force,2 18% of the total sample was considered overweight and 10% was obese. At baseline, on average, participants from the intervention and control groups reported a daily consumption of 2.3 portions of fruits and vegetables (SD, 1.0 portions), daily physical activity of 1.9 hours (SD, 1.3 hours), and daily screen time of 1.4 hours (SD, 1.1 hours). Fruit and vegetable consumption was significantly different between groups at baseline: On average, participants in the control group consumed 0.7 daily portions of fruits and vegetables more than in participants the intervention group (t[47] ¼ 2.38; P < .05).
Monitoring Adherence On average, children reported the 3 target behaviors (number of portions of fruits or vegetables, steps, and minutes of screen time) 61% of the days (mean, 34.2; median, 31.5; SD, 13.3 days). In the first 4 weeks, on average, participants sent more than 0.75 daily messages (eg, in the first week 0.94 messages were sent on average). However, in weeks 5 and 6 the average weekly SMS sent decreased to 0.65, and in weeks 7 and 8 to 0.45.
Satisfaction with Monitoring and Feedback Program A significant number of children (95%) were satisfied with the program and would recommend it to a friend (75%). Children enjoyed sending the monitoring SMS (71.4%) and judged the feedback messages to be adequate (90.5%). All children enjoyed using the pedometer.
Preliminary Efficacy of Monitoring and Feedback Program Table 2 gives an overview of data for the 3 target behaviors at pre- and postintervention and follow-up. Analyses of covariance with gender and baseline levels of the 3 target behaviors as covariates were used. For fruit and vegetable consumption, the
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015
78 Fassnacht et al
Table 1. Demographics and Baseline Behavior of Total Sample of Elementary Schoolchildren in Portugal Overall (N ¼ 49) 26 (53.1)
Intervention (n ¼ 22) 8 (36.4)
Control (n ¼ 27) 18 (66.7)
P *
Age, y (mean [SD])b
9.6 (0.4)
9.5 (0.3)
9.6 (0.4)
NS
Body mass index z-scores (mean [SD])c
0.8 (1.1)
1.0 (1.3)
0.6 (0.9)
NS
Body mass index z-score intervals (n [%]) Normal weight (z # 1.0) Overweight (1.0 < z # 2.0) Obese (z > 2.0)
35 (71.4) 9 (18.4) 5 (10.2)
14 (63.6) 4 (18.2) 4 (18.2)
21 (77.8) 5 (18.5) 1 (3.7)
Fruit and vegetables (mean [SD])d
2.3 (1.0)
1.9 (0.8)
2.6 (1.1)
*
1.9 (1.3)
1.7 (1.1)
2.1 (1.4)
NS
1.4 (1.1)
1.2 (1.0)
1.5 (1.2)
NS
Female (n [%])a
Physical activity (mean [SD])
e
Screen time (mean [SD])f
NS indicates not significant. a Chi-square test, c2(1) ¼ 4.47, P ¼ .03; bIndependent samples t test, t(47) ¼ 1.57, P ¼ .12; cIndependent samples t test, t(47) ¼ 1.09, P ¼ .28; d0 ¼ 0 portions, 5 ¼ $5 portions. Independent samples t test, t(47) ¼ 2.83, P ¼ .02; e0 ¼ 0 hours, 5 ¼ $5 hours. Independent samples t test, t(43) ¼ 1.11, P ¼ .27; f0 ¼ 0 hours, 5 ¼ $5 hours. Independent samples t test, t(47) ¼ 0.92, P ¼ .36; *P < .05. Note: All results are based on children. Only the intervention group received the monitoring and feedback program. results showed that there was a statistically significant main effect of time (F[1.91,78.46] ¼ 3.98; P < .05; partial h2 ¼ 0.09) and a significant interaction effect of time group (F [1.91,78.46] ¼ 3.48; P < .05; partial h2 ¼ 0.08) after controlling for gender and baseline levels of fruit and vegetable consumption. In addition, the interaction effects of the covariates and time are reported: The interaction effect of time gender (covariate) did not reach statistical significance (F[1.91,78.46] ¼ 0.38; P ¼ .68; partial h2 ¼ 0.009). However, the interaction effect of time baseline consumption of fruits and vegetables (covariate) was
statistically significant (F[1.91,78.46] ¼ 8.21; P < .01; partial h2 ¼ 0.17). For physical activity, there was a significant main effect of time (F[2,60] ¼ 9.94; P < .01; partial h2 ¼ .41) after controlling for gender and baseline levels of physical activity. However, the interaction effect of time group did not reach statistical significance (F[2,60] ¼ 1.48; P ¼ .24; partial h2 ¼ 0.09). The interaction effect of time gender (covariate) did not reach statistical significance (F[2,60] ¼ 1.73; P ¼ .19; partial h2 ¼ 0.06). However, the interaction effect of time baseline physical activity levels (covariate) was statistically sig-
Table 2. Impact of Text Messaging on Target Behaviors (Baseline, Postintervention, and Follow-up) Baseline
Postintervention
Follow-up
n Mean SD Fruit and vegetables Intervention 22 2.0 (0.8) Control 27 2.6 (1.1)
n
Mean
SD
n
Mean
SD
22 26
2.8 2.3
(1.1) (1.3)
21 24
2.5 2.3
(1.2) (1.3)
Physical activity Intervention Control
20 25
1.7 2.1
(1.1) (1.4)
21 23
1.6 1.6
(0.9) (1.0)
19 23
1.9 1.7
(1.2) (1.2)
Screen time Intervention Control
22 27
1.2 1.6
(0.9) (1.2)
20 24
0.9 1.1
(0.6) (0.8)
19 23
0.7 1.6
(0.5) (1.1)
Note: Baseline indicates pre-intervention scores; postintervention, scores after 8 weeks (no) monitoring; follow-up, scores 4 weeks after (no) monitoring.
nificant (F[2,60] ¼ 29.40; P < .001; partial h2 ¼ 0.50). For screen time, results showed that the main effect of time did not reach statistical significance (F[1.92,65.32] ¼ 2.61; P ¼ .08; partial h2 ¼ 0.07). However, there was a significant interaction effect of time group (F[1.92,65.32] ¼ 4.48; P < .05; partial h2 ¼ 0.12) after controlling for gender and baseline levels of screen time. The interaction effect of time by gender (covariate) did not reach statistical significance (F[1.91,78.46] ¼ 0.02; P ¼ .97; partial h2 ¼ 0.001). However, the interaction effect of time baseline screen time (covariate) was statistically significant (F[1.91,78.46] ¼ 32.92; P < .001; partial h2 ¼ 0.49).
DISCUSSION The current pilot study aimed to assess adherence, satisfaction, and preliminary efficacy of an SMS-based monitoring and feedback system regarding 3 target behaviors (ie, fruit and vegetable consumption, physical activity, screen time) of 8- to 10-yearold children. Results of the current study indicated good adherence to the SMS intervention; children submitted 61% of the required monitoring SMS. This finding is consistent with results of previous research13 indicating that children adhere well
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 to technology-enhanced systems.6 Comparing text messaging (43% completed) and paper diaries (19% completed), Shapiro et al6 reported that text messaging enhances compliance to self–monitoring; thus, children seem to prefer technological tailored programs. The present study also confirms findings of Bauer et al,13 who found a good adherence rate (67%) with the SMS intervention. Adherence and acceptability of the self-monitoring program might also be related to the supportive positive reinforcement, an important element of social cognitive theory. Because mobile phones and text messaging are used globally as communication tools, this appears to be a promising approach to support individuals' efforts to improve healthrelated behaviors. A recent review of mobile text messaging use in clinical and health behavior interventions revealed that SMS interventions received promising acceptance and efficacy rates.25 However, whereas the average weekly number of SMS sent was acceptable from weeks 1 to 6, it decreased in the last 2 weeks of the intervention to less than any other day (0.45). Possible strategies for keeping children engaged include sending motivational messages in the last 2 weeks (eg, ‘‘Only 2 weeks to go, you’re almost there!’’) and further classroom visits to engage children to complete the program. Preliminary results of the efficacy indicate that behavioral improvement for fruit and vegetable consumption and screen time could be achieved. Participants in the intervention group consumed significantly more fruits and vegetables over time compared with the control group. Regarding screen time, participants from the intervention group spent significantly less time in front of the screen (television, video games, computers, etc) over time compared with the control group. Hence, the anticipated objectives for these 2 behaviors were obtained. In line with the current study, Epstein et al26 observed that reinforcement decreased sedentary behavior and increased fruit and vegetable consumption in obese children. Interestingly, the hypothesis regarding an increase in physical activity, in which the objective was $ 1 h/d of moderate physical activity, could not be retained. However, on a
group level, participants from both groups had achieved the anticipated goal before the intervention. Thus, children received a reinforcing feedback message if they did not reach the expected goal of 10,000 steps/d. However, in approximately 75% of the sent messages, children reported $ 10,000 steps and received rewarding feedback for the achieved goal. Setting the goal to 15,000 steps, for example, would have resulted in only 20% of achieved goals in the present study. Hence, children would have received reinforcing messages more frequently and might have tried harder to achieve the goal. It thus seems crucial that the objective for physical activity (10,000 steps/d in the current study) will be adjusted in future studies (eg, 15,000 steps/d) to investigate the question of whether the intervention group would increase activity levels with reinforcing feedback. A limitation of the current study was its small sample size. A further limitation was that children were randomized class-wise rather than individually. It is not surprising that baseline differences in the 3 target behaviors occurred. Differences in fruit and vegetable consumption were found, with the control group reporting a significant higher daily consumption. In addition, gender was not equally distributed across the 2 groups. Although the methodologically sound approach would have been to randomize by individual participants, this could have compromised the results because children talk about their participation during class. To control for any influence of gender and baseline levels of the 3 target behaviors, these variables were included as covariates in the analyses. The sample size of this pilot study was too small to ultimately test the efficacy of the intervention program. Future research should randomize at the class level in an adequately powered cluster randomized design. Another limitation is the design of the study was that it did not control for the effect of self-monitoring alone, and simply self-monitoring can lead to behavior change in healthy eating and physical activity.27 However, for this pilot study only a monitoring plus SMS feedback group was compared with a no-monitoring
Fassnacht et al 79 group. The goal of this study was to establish feasibility and adherence first. Because of the limited methodological design, the preliminary efficacy of the SMS feedback for fruit and vegetable consumption and screen time could not be interpreted separately from the effect of simply self-monitoring. To parse out the effect of the SMS feedback program, a larger randomized, controlled efficacy trial could include 3 groups: SMS monitoring plus feedback, paper–pencil monitoring without feedback, and no monitoring. The results indicate that participation in the SMS program may be associated with positive behavioral changes and encourage systematic study of the efficacy of the intervention in a largescale, randomized, controlled trial.
IMPLICATIONS FOR RESEARCH AND PRACTICE In past years information and communication technology has become an effective way to provide behavior change interventions.4,25,28 Hence, it is necessary to concentrate on the evaluation of existing programs or the development of adapted versions.12 Despite the need for a larger randomized, controlled efficacy trial, future studies should also investigate the program in a clinical context. In addition to promoting health behavior in normal weight populations, the SMS-based monitoring and feedback system could also be used as an add-on to regular intervention programs for overweight and obese children. In this study, almost one-third of children were overweight (18.4%) or obese (10.2%), which shows a need for tailored intervention programs for these individuals. Thus, adherence of overweight or obese children to the SMS program could be explored. Target behaviors and cutoff levels could be adjusted for this specific group of individuals and a randomized controlled trial could identify whether behavioral changes regarding health behavior might occur. Because previous research has shown that selfmonitoring leads to increased adherence to treatment goals, this could result in improved outcomes,12 and
80 Fassnacht et al overweight and obese children in particular could benefit from such programs.
ACKNOWLEDGMENTS This research was supported by a Fundac¸~ao para a Ci^encia e a Tecnologia/Foundation for Science and Technology, Portugal Research Grant (PTDC/PSI-PCL/099981/2008) to Dr Machado. The SMS program was developed at the Center for Psychotherapy Research, University Hospital Heidelberg (Germany). The authors thank Drs Stephanie Bauer and Markus Moessner for advisory support with the program, Margarida Silva and Diana Costa for support with the educational sessions, and Lutfi Arikan for technical assistance.
REFERENCES 1. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Geneva, Switzerland: World Health Organization; 2000. 2. Cole TJ, Lobstein T. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity. Pediatr Obes. 2012;7:284-294. 3. Webb TL, Joseph J, Yardley L, Michie S. Using the internet to promote health behavior change: a systematic review and meta-analysis of the impact of theoretical basis, use of behavior change techniques, and mode of delivery on efficacy. J Med Internet Res. 2010;12:e4. 4. Nguyen B, Kornman KP, Baur LA. A review of electronic interventions for prevention and treatment of overweight and obesity in young people. Obes Rev. 2011;12:e298-e314. Bodurtha J, 5. Smith AJ, Skow A, Kinra S. Health information technology in screening and treatment of child obesity: a systematic review. Pediatrics. 2013;131:e894-e902. 6. Shapiro JR, Bauer S, Hamer RM, Kordy H, Ward D, Bulik CM. Use of text messaging for monitoring sugarsweetened beverages, physical activity, and screen time in children: a pilot study. J Nutr Educ Behav. 2008;40:385-391.
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 7. Ferrer-Roca O, Cardenas A, DiazCardama A, Pulido P. Mobile phone text messaging in the management of diabetes. J Telemed Telecare. 2004;10: 282-285. 8. Anhj J, Mldrup C. Feasibility of collecting diary data from asthma patients through mobile phones and SMS (short message service): response rate analysis and focus group evaluation from a pilot study. J Med Internet Res. 2004;6:e42. 9. Rodgers A, Corbett T, Bramley D, et al. Do u smoke after txt? Results of a randomised trial of smoking cessation using mobile phone text messaging. Tob Control. 2005;14:255-261. 10. Kauer DS, Reid CS, Crooke DAH, et al. Self-monitoring using mobile phones in the early stages of adolescent depression: randomized controlled trial. J Med Internet Res. 2012;14:e67. 11. Bauer S, Okon E, Meermann R, Kordy H. Technology-enhanced maintenance of treatment gains in eating disorders: efficacy of an intervention delivered via text messaging. J Consult Clin Psychol. 2012;80:700-706. 12. de Niet J, Timman R, Bauer S, et al. The effect of a short message service maintenance treatment on body mass index and psychological well-being in overweight and obese children: a randomized controlled trial. Pediatr Obes. 2012;7:205-219. 13. Bauer S, de Niet J, Timman R, Kordy H. Enhancement of care through self-monitoring and tailored feedback via text messaging and their use in the treatment of childhood overweight. Patient Educ Couns. 2010;79:315-319. 14. Cardoso G, Gomes MdC, Espanha R, Lapa T. E-Generation: Os Usos de Media Pelas Crianc¸as e Jovens em Portugal. [E-Generation: The Use of Media by Children and Adolescents in Portugal]. Lisbon, Portugal: CIES/ISCTE; 2007. 15. Bauer S, Percevic R, Okon E, Meermann R, Kordy H. Use of text messaging in the aftercare of patients with bulimia nervosa. Eur Eat Disord Rev. 2003;11:279-290. 16. Bandura A. Social cognitive theory. In: Vasta R, ed. Annals of Child Development. Greenwich, CT: JAI Press; 1989:1-60. 17. Currie C, Gabhainn SN, Godeau E, et al. Inequalities in children’s health: HBSC,
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
International Report from the 2005/2006 Survey. Geneva, Switzerland: World Health Organization; 2008:83-109. Vereecken CA, Maes LA. Belgian study on the reliability and relative validity of the Health Behaviour in School-Aged Children food-frequency questionnaire. Public Health Nutr. 2003;6: 581-588. Golan M, Weizman A. Reliability and validity of the Family Eating and Activity Habits Questionnaire. Eur J Clin Nutr. 1998;52:771-777. Cole TJ. A chart to link child centiles of body mass index, weight and height. Eur J Clin Nutr. 2002;56:1194-1199. Tudor-Locke C, Pangrazi RP, Corbin CB, et al. BMI-referenced standards for recommended pedometer-determined steps/day in children. Prev Med. 2004;38: 857-864. Tudor-Locke C, Hatano Y, Pangrazi RP, Kang M. Revisiting ‘‘how many steps are enough?’’. Med Sci Sports Exerc. 2008;40(7 suppl): S537-S543. World Health Organization. Fruit and Vegetables Promotion Initiative. Geneva, Switzerland: World Health Organization; 2003. Conover WJ. Practical Nonparametric Statistics. 3rd ed. New York, NY: Wiley; 1999. Wei J, Hollin I, Kachnowski S. A review of the use of mobile phone text messaging in clinical and healthy behaviour interventions. J Telemed Telecare. 2011;17:41-48. Epstein LH, Paluch RA, Kilanowski CK, Raynor HA. The effect of reinforcement or stimulus control to reduce sedentary behavior in the treatment of pediatric obesity. Health Psychol. 2004;23:371-380. Williams SL, French DP. What are the most effective intervention techniques for changing physical activity selfefficacy and physical activity behaviour—and are they the same? Health Educ Res; 2011:308-322. Lau PWC, Lau EY, Wong DP, Ransdell L. A systematic review of information and communication technology-based interventions for promoting physical activity behavior change in children and adolescents. J Med Internet Res. 2011;13:e48.
INTRODUCTION The rise in childhood obesity has been drastic over the past decades and is one of the most serious public health challenges of the 21st century.1 According to the International Association for the Study of Obesity, over 200 million school-aged children are overweight, which makes this generation the first predicted to have a shorter lifespan than their parents.2 Childhood obesity is mainly associated with unhealthy diets and low physical activity. Health promotion and intervention programs for childhood obesity need to support and facilitate an increase in physical activity and healthier diets. The use of new information and communication technologies has
shown promising results in promoting health behaviors,3 preventing childhood obesity,4 and maintaining weight loss in overweight children.5 For example, the short message service (SMS) has proven to be useful in promoting health behaviors in children.6 In fact, SMS has been acceptable for providing support, effecting behavior change, and/or maintaining treatment gains in diabetes,7 asthma,8 smoking cessation,9 depressive symptoms,10 bulimia nervosa,11 and childhood obesity.12 Children prefer novel devices, and technology-enhanced systems showed positive effects on participants' adherence rates.6 Furthermore, SMS has several advantages such as accessibility at any time, fast and instant information, guidance and
1
Department of Psychology, James Cook University, Singapore Psychotherapy and Psychopathology Research Unit, CIPsi, School of Psychology, University of Minho, Braga, Portugal 3 Centre for Mental Health Research, Australian National University, Canberra School of Psychology, Campus de Gualtar, Braga, Portugal Address for correspondence: Paulo P. P. Machado, PhD, School of Psychology, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal; Phone: (þ351) 253-604220; Fax: (þ351) 253-604224; E-mail: [email protected] Ó2015 SOCIETY FOR NUTRITION EDUCATION AND BEHAVIOR http://dx.doi.org/10.1016/j.jneb.2014.08.006 2
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015
advice without major effort (after messages are created), and low costs.13 The use of Internet and mobile phones has increased in the past 2 decades; according to Cardoso et al,14 84.2% of children aged 9–12 years in Portugal own a mobile phone and 88% use it to type messages. Bauer et al15 developed an Internetbased computer program used in previous research to provide support and promote behavior change via SMS. Methods from social cognitive theory and behavioral models were considered and used in developing the program. Self-monitoring and immediate feedback based on specific goals are important elements in behavioral theory. The idea that health behavior will change with specific goals when positive reinforcement is provided is based on social cognitive theory.16 Shapiro et al6 used an adopted version of the program developed by Bauer et al15 and showed that text messaging might be a useful tool for self-monitoring sugar-sweetened beverages, physical activity, and screen time. Children monitored target behaviors via SMS with feedback or via paper diaries, or participated in a no-monitoring condition. In the current study, the researchers
75
76 Fassnacht et al adapted this program to promote health behaviors (eg, fruit and vegetable consumption, physical activity, screen time) in children. Compared with the study by Shapiro et al, children were randomized by classes into SMS monitoring vs no monitoring.6 The intervention consisted of an interaction between participants and provider via SMS. Participants sent text messages at regular intervals and received supportive feedback messages based on their entries. The current pilot study aimed to explore participants' adherence to and satisfaction with the SMS-based monitoring and feedback system. A secondary aim of the pilot study was to explore the preliminary efficacy of the program to promote health behaviors. To test this aim, researchers hypothesized that the intervention group would report increased fruit and vegetable consumption and physical activity and decreased screen time.
METHODS Children from the fourth grade were recruited from an elementary school in Braga, Portugal. By tossing a coin, the children of 2 school classes were assigned to either a monitoring (intervention: n ¼ 22) or control (n ¼ 27) condition. Children (n ¼ 49) between the ages of 8 and 10 years (mean, 9.6 years; SD, 0.4 years) were included in the study regardless of weight or ethnicity. The intervention with the SMS-based monitoring and feedback system took place over 8 weeks.
Instruments Participants in both groups completed a self-report health behavior questionnaire at the beginning of the first educational group session (baseline), after 8 weeks of monitoring (postintervention), and at 4 weeks' followup in class. The researchers measured usual intake of fruits and vegetables using questions of a paper-based food frequency questionnaire based on the Health Behaviour in SchoolAged Children Study.17 Children were asked, ‘‘How many times a week do you usually eat/drink .’’ followed by a list of food and beverage items including fruits and vegetables. Options were never, less than once a
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 week, once a week, 2–4 d/wk, 5–6 d/wk, once every day, and several times every day. The food frequency questionnaire showed good reliability and validity.18 Daily physical activity and screen time were measured using 2 questions from the Family Eating and Activity Habits Questionnaire19: ‘‘How many hours per day on average did you participate in activities such as fast walking, swimming, ball games?’’; ‘‘How many minutes did you spend in front of the screen (eg, television, computer, video games)?’’ The Family Eating and Activity Habits Questionnaire showed good reliability and validity.19 These questions were used to explore preliminary efficacy. All questionnaires were translated from English into Portuguese and back-translated by the authors. Items were carefully checked by native speakers and changed or edited when necessary. Height and weight were assessed (baseline) without shoes using a digital scale and a stadiometer. Body mass index (kg/m2) for age was calculated for children according to the International Obesity Task Force using SD scores (z-scores) as recommended by Cole.2,20 Children in the intervention group received a pedometer (Silva Pedometer Plus, Silva AB, Bromma, Sweden, 2007) to monitor their steps per day accurately. In addition, children from the intervention group completed a satisfaction questionnaire with 8 questions (eg, ‘‘How much fun did you have doing this program?’’; response options were No fun at all, No fun, Neither . nor, Fun, and Lots of fun) after the intervention to explore children's overall satisfaction with specific components of the program (eg, appropriateness of feedback messages).
Procedure Parents and children received information sheets with necessary details on the planned program (ie, use of mobile phones and pedometers for the intervention group), including informed consent. If parents agreed to their child's participation, they were invited to participate in 1 educational session for parents, were informed about the importance of health behaviors, and were given
detailed information about the program. All children participated in 2 60-minute educational sessions presented in a group format and facilitated by 2 trained psychologists. The only material that differed between the intervention and control groups was presented in session 2, which included detailed information about and training with the SMS program. Session 1 focused on increasing physical activity, decreasing screen time, and the risks of sedentary behavior. Session 2 focused on a healthy diet in general and the importance of the consumption of fruits and vegetables specifically. Children from the intervention group were asked to monitor their fruit and vegetable consumption, physical activity, and screen time daily. Regarding fruit and vegetable consumption, specific serving sizes were discussed and children were asked to report the exact amount of daily fruit and vegetable intake. Children were instructed to report data in a standard format via SMS. Specific questions to obtain this information included ‘‘How many fruits and vegetables did you eat today?’’ ‘‘How many steps did you achieve with the pedometer today?’’ and ‘‘How many minutes did you spend in front of the screen today?’’ To create normative values for the 3 behaviors, the authors used previous research to define 5 portions of fruits and vegetables, healthy amounts of physical activity at 10,000 steps/d, and < 60 min/d screen time. Recommendations for physical activity and sedentary behavior were based on Tudor-Locke et al21,22 and Shapiro et al,6 because national recommendations do not exist specifically for Portugal. Recommendations for fruit and vegetable consumption were based on the World Health Organization, which considers an ‘‘adequate quantity’’ as at least 400–500 g/d, which is equivalent to 5 servings of 80 g of fruits and/or vegetables.23 Children were asked to use their parent's mobile phones, because parents were to support and help the children throughout the program. Both children and parents became acquainted with the appropriate use of the pedometer and how to transfer the data to the daily SMS to submit the SMS together. All costs for mobile
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 phone usage regarding the program were reimbursed; however, no incentives were provided. Messages from participants were sent to a modem connected to a secure Web server. Incoming and outgoing messages were recorded and the software program automatically suggested tailored feedback messages. For the current study, fruit and vegetable consumption replaced consumption of sugar-sweetened beverages. A pool of 900 existing messages from previous research with the SMS program was translated from English into Portuguese and new ones were created to address fruit and vegetable consumption.6 Algorithms for the feedback messages were based on how many of the 3 goals were met and improvement or deterioration from the previous day (eg, ‘‘Great, you met your goal for physical activity and screen time! What happened to fruits and vegetables?’’). The prompted answers were transformed dichotomously by the software program according to whether the behavioral goal was achieved and how this outcome related to the previous status report of the child. Details about the algorithm used and the development of the program were described elsewhere.6,15 The feedback messages aimed to motivate and encourage children to reach each behavioral goal and to support and reinforce positive development based on improved or deteriorated behavior. Giving technological constraints, feedback messages had a limit of 160 characters. The Ethics Committee of the School of Psychology approved the study protocol and all parents/legal guardians signed informed consent after child assent.
time as the within-subjects factor and condition as the betweensubjects factor to compare the amount of consumed fruits and vegetables, physical activity, and screen time between conditions (intervention and control group) at pre- and postintervention and follow-up (time). Gender and baseline levels of the 3 target behaviors were included as covariates. Main effects of time and interaction effects of time group were found. In addition, interaction effects of the covariates (gender, baseline levels of the 3 behaviors) and time were observed. Assumptions of sphericity were tested using the Mauchly sphericity test. For fruit and vegetable consumption (c2 [2] ¼ 7.27; P < .05) and screen time (c2 [2] ¼ 6.79; P < .05), Mauchly tests indicated that the assumption of sphericity had been violated; therefore, degrees of freedom were corrected using Huynh–Feldt estimates of sphericity (ε ¼ .96). Assumptions of normality were tested using the Shapiro–Wilks test. Because of the small sample size, it was not surprising that the assumption for physical activity and screen time was violated for both the control and intervention groups. To test the influence of the normality violation, the data of these 2 variables were both z- and rank-transformed and subsequent ANCOVAs were performed on the transformed data. The 2 procedures gave nearly identical results; hence, the assumptions underlying the ANCOVA are likely to be reasonable.24 In the following only the parametric results of the ANCOVAs are reported. All analyses were performed using SPSS (version 20.0, SPSS Inc, Chicago, IL, 2011).
Data Analysis
RESULTS Participants
The authors measured adherence to the monitoring system by dividing the total number of required monitoring assessments (ie, 56 days) by the number of days participants actually sent their monitored behavior via SMS. To explore the preliminary efficacy of the monitoring and feedback system, the researchers conduced repeated-measures ANCOVA with
Table 1 lists demographic variables and baseline behaviors (fruit and vegetable consumption, physical activity, and screen time) of the sample. A total of 51 children were eligible and 49 participated in the study (2 did not provide signed parental informed consent). The intervention group consisted of significantly more males (n ¼ 14) compared with the control group (n ¼ 9). Children were aged 8
Fassnacht et al 77 to 10 years (mean, 9.6; SD, 0.4). Using body mass index cutoffs provided by the International Obesity Task Force,2 18% of the total sample was considered overweight and 10% was obese. At baseline, on average, participants from the intervention and control groups reported a daily consumption of 2.3 portions of fruits and vegetables (SD, 1.0 portions), daily physical activity of 1.9 hours (SD, 1.3 hours), and daily screen time of 1.4 hours (SD, 1.1 hours). Fruit and vegetable consumption was significantly different between groups at baseline: On average, participants in the control group consumed 0.7 daily portions of fruits and vegetables more than in participants the intervention group (t[47] ¼ 2.38; P < .05).
Monitoring Adherence On average, children reported the 3 target behaviors (number of portions of fruits or vegetables, steps, and minutes of screen time) 61% of the days (mean, 34.2; median, 31.5; SD, 13.3 days). In the first 4 weeks, on average, participants sent more than 0.75 daily messages (eg, in the first week 0.94 messages were sent on average). However, in weeks 5 and 6 the average weekly SMS sent decreased to 0.65, and in weeks 7 and 8 to 0.45.
Satisfaction with Monitoring and Feedback Program A significant number of children (95%) were satisfied with the program and would recommend it to a friend (75%). Children enjoyed sending the monitoring SMS (71.4%) and judged the feedback messages to be adequate (90.5%). All children enjoyed using the pedometer.
Preliminary Efficacy of Monitoring and Feedback Program Table 2 gives an overview of data for the 3 target behaviors at pre- and postintervention and follow-up. Analyses of covariance with gender and baseline levels of the 3 target behaviors as covariates were used. For fruit and vegetable consumption, the
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015
78 Fassnacht et al
Table 1. Demographics and Baseline Behavior of Total Sample of Elementary Schoolchildren in Portugal Overall (N ¼ 49) 26 (53.1)
Intervention (n ¼ 22) 8 (36.4)
Control (n ¼ 27) 18 (66.7)
P *
Age, y (mean [SD])b
9.6 (0.4)
9.5 (0.3)
9.6 (0.4)
NS
Body mass index z-scores (mean [SD])c
0.8 (1.1)
1.0 (1.3)
0.6 (0.9)
NS
Body mass index z-score intervals (n [%]) Normal weight (z # 1.0) Overweight (1.0 < z # 2.0) Obese (z > 2.0)
35 (71.4) 9 (18.4) 5 (10.2)
14 (63.6) 4 (18.2) 4 (18.2)
21 (77.8) 5 (18.5) 1 (3.7)
Fruit and vegetables (mean [SD])d
2.3 (1.0)
1.9 (0.8)
2.6 (1.1)
*
1.9 (1.3)
1.7 (1.1)
2.1 (1.4)
NS
1.4 (1.1)
1.2 (1.0)
1.5 (1.2)
NS
Female (n [%])a
Physical activity (mean [SD])
e
Screen time (mean [SD])f
NS indicates not significant. a Chi-square test, c2(1) ¼ 4.47, P ¼ .03; bIndependent samples t test, t(47) ¼ 1.57, P ¼ .12; cIndependent samples t test, t(47) ¼ 1.09, P ¼ .28; d0 ¼ 0 portions, 5 ¼ $5 portions. Independent samples t test, t(47) ¼ 2.83, P ¼ .02; e0 ¼ 0 hours, 5 ¼ $5 hours. Independent samples t test, t(43) ¼ 1.11, P ¼ .27; f0 ¼ 0 hours, 5 ¼ $5 hours. Independent samples t test, t(47) ¼ 0.92, P ¼ .36; *P < .05. Note: All results are based on children. Only the intervention group received the monitoring and feedback program. results showed that there was a statistically significant main effect of time (F[1.91,78.46] ¼ 3.98; P < .05; partial h2 ¼ 0.09) and a significant interaction effect of time group (F [1.91,78.46] ¼ 3.48; P < .05; partial h2 ¼ 0.08) after controlling for gender and baseline levels of fruit and vegetable consumption. In addition, the interaction effects of the covariates and time are reported: The interaction effect of time gender (covariate) did not reach statistical significance (F[1.91,78.46] ¼ 0.38; P ¼ .68; partial h2 ¼ 0.009). However, the interaction effect of time baseline consumption of fruits and vegetables (covariate) was
statistically significant (F[1.91,78.46] ¼ 8.21; P < .01; partial h2 ¼ 0.17). For physical activity, there was a significant main effect of time (F[2,60] ¼ 9.94; P < .01; partial h2 ¼ .41) after controlling for gender and baseline levels of physical activity. However, the interaction effect of time group did not reach statistical significance (F[2,60] ¼ 1.48; P ¼ .24; partial h2 ¼ 0.09). The interaction effect of time gender (covariate) did not reach statistical significance (F[2,60] ¼ 1.73; P ¼ .19; partial h2 ¼ 0.06). However, the interaction effect of time baseline physical activity levels (covariate) was statistically sig-
Table 2. Impact of Text Messaging on Target Behaviors (Baseline, Postintervention, and Follow-up) Baseline
Postintervention
Follow-up
n Mean SD Fruit and vegetables Intervention 22 2.0 (0.8) Control 27 2.6 (1.1)
n
Mean
SD
n
Mean
SD
22 26
2.8 2.3
(1.1) (1.3)
21 24
2.5 2.3
(1.2) (1.3)
Physical activity Intervention Control
20 25
1.7 2.1
(1.1) (1.4)
21 23
1.6 1.6
(0.9) (1.0)
19 23
1.9 1.7
(1.2) (1.2)
Screen time Intervention Control
22 27
1.2 1.6
(0.9) (1.2)
20 24
0.9 1.1
(0.6) (0.8)
19 23
0.7 1.6
(0.5) (1.1)
Note: Baseline indicates pre-intervention scores; postintervention, scores after 8 weeks (no) monitoring; follow-up, scores 4 weeks after (no) monitoring.
nificant (F[2,60] ¼ 29.40; P < .001; partial h2 ¼ 0.50). For screen time, results showed that the main effect of time did not reach statistical significance (F[1.92,65.32] ¼ 2.61; P ¼ .08; partial h2 ¼ 0.07). However, there was a significant interaction effect of time group (F[1.92,65.32] ¼ 4.48; P < .05; partial h2 ¼ 0.12) after controlling for gender and baseline levels of screen time. The interaction effect of time by gender (covariate) did not reach statistical significance (F[1.91,78.46] ¼ 0.02; P ¼ .97; partial h2 ¼ 0.001). However, the interaction effect of time baseline screen time (covariate) was statistically significant (F[1.91,78.46] ¼ 32.92; P < .001; partial h2 ¼ 0.49).
DISCUSSION The current pilot study aimed to assess adherence, satisfaction, and preliminary efficacy of an SMS-based monitoring and feedback system regarding 3 target behaviors (ie, fruit and vegetable consumption, physical activity, screen time) of 8- to 10-yearold children. Results of the current study indicated good adherence to the SMS intervention; children submitted 61% of the required monitoring SMS. This finding is consistent with results of previous research13 indicating that children adhere well
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 to technology-enhanced systems.6 Comparing text messaging (43% completed) and paper diaries (19% completed), Shapiro et al6 reported that text messaging enhances compliance to self–monitoring; thus, children seem to prefer technological tailored programs. The present study also confirms findings of Bauer et al,13 who found a good adherence rate (67%) with the SMS intervention. Adherence and acceptability of the self-monitoring program might also be related to the supportive positive reinforcement, an important element of social cognitive theory. Because mobile phones and text messaging are used globally as communication tools, this appears to be a promising approach to support individuals' efforts to improve healthrelated behaviors. A recent review of mobile text messaging use in clinical and health behavior interventions revealed that SMS interventions received promising acceptance and efficacy rates.25 However, whereas the average weekly number of SMS sent was acceptable from weeks 1 to 6, it decreased in the last 2 weeks of the intervention to less than any other day (0.45). Possible strategies for keeping children engaged include sending motivational messages in the last 2 weeks (eg, ‘‘Only 2 weeks to go, you’re almost there!’’) and further classroom visits to engage children to complete the program. Preliminary results of the efficacy indicate that behavioral improvement for fruit and vegetable consumption and screen time could be achieved. Participants in the intervention group consumed significantly more fruits and vegetables over time compared with the control group. Regarding screen time, participants from the intervention group spent significantly less time in front of the screen (television, video games, computers, etc) over time compared with the control group. Hence, the anticipated objectives for these 2 behaviors were obtained. In line with the current study, Epstein et al26 observed that reinforcement decreased sedentary behavior and increased fruit and vegetable consumption in obese children. Interestingly, the hypothesis regarding an increase in physical activity, in which the objective was $ 1 h/d of moderate physical activity, could not be retained. However, on a
group level, participants from both groups had achieved the anticipated goal before the intervention. Thus, children received a reinforcing feedback message if they did not reach the expected goal of 10,000 steps/d. However, in approximately 75% of the sent messages, children reported $ 10,000 steps and received rewarding feedback for the achieved goal. Setting the goal to 15,000 steps, for example, would have resulted in only 20% of achieved goals in the present study. Hence, children would have received reinforcing messages more frequently and might have tried harder to achieve the goal. It thus seems crucial that the objective for physical activity (10,000 steps/d in the current study) will be adjusted in future studies (eg, 15,000 steps/d) to investigate the question of whether the intervention group would increase activity levels with reinforcing feedback. A limitation of the current study was its small sample size. A further limitation was that children were randomized class-wise rather than individually. It is not surprising that baseline differences in the 3 target behaviors occurred. Differences in fruit and vegetable consumption were found, with the control group reporting a significant higher daily consumption. In addition, gender was not equally distributed across the 2 groups. Although the methodologically sound approach would have been to randomize by individual participants, this could have compromised the results because children talk about their participation during class. To control for any influence of gender and baseline levels of the 3 target behaviors, these variables were included as covariates in the analyses. The sample size of this pilot study was too small to ultimately test the efficacy of the intervention program. Future research should randomize at the class level in an adequately powered cluster randomized design. Another limitation is the design of the study was that it did not control for the effect of self-monitoring alone, and simply self-monitoring can lead to behavior change in healthy eating and physical activity.27 However, for this pilot study only a monitoring plus SMS feedback group was compared with a no-monitoring
Fassnacht et al 79 group. The goal of this study was to establish feasibility and adherence first. Because of the limited methodological design, the preliminary efficacy of the SMS feedback for fruit and vegetable consumption and screen time could not be interpreted separately from the effect of simply self-monitoring. To parse out the effect of the SMS feedback program, a larger randomized, controlled efficacy trial could include 3 groups: SMS monitoring plus feedback, paper–pencil monitoring without feedback, and no monitoring. The results indicate that participation in the SMS program may be associated with positive behavioral changes and encourage systematic study of the efficacy of the intervention in a largescale, randomized, controlled trial.
IMPLICATIONS FOR RESEARCH AND PRACTICE In past years information and communication technology has become an effective way to provide behavior change interventions.4,25,28 Hence, it is necessary to concentrate on the evaluation of existing programs or the development of adapted versions.12 Despite the need for a larger randomized, controlled efficacy trial, future studies should also investigate the program in a clinical context. In addition to promoting health behavior in normal weight populations, the SMS-based monitoring and feedback system could also be used as an add-on to regular intervention programs for overweight and obese children. In this study, almost one-third of children were overweight (18.4%) or obese (10.2%), which shows a need for tailored intervention programs for these individuals. Thus, adherence of overweight or obese children to the SMS program could be explored. Target behaviors and cutoff levels could be adjusted for this specific group of individuals and a randomized controlled trial could identify whether behavioral changes regarding health behavior might occur. Because previous research has shown that selfmonitoring leads to increased adherence to treatment goals, this could result in improved outcomes,12 and
80 Fassnacht et al overweight and obese children in particular could benefit from such programs.
ACKNOWLEDGMENTS This research was supported by a Fundac¸~ao para a Ci^encia e a Tecnologia/Foundation for Science and Technology, Portugal Research Grant (PTDC/PSI-PCL/099981/2008) to Dr Machado. The SMS program was developed at the Center for Psychotherapy Research, University Hospital Heidelberg (Germany). The authors thank Drs Stephanie Bauer and Markus Moessner for advisory support with the program, Margarida Silva and Diana Costa for support with the educational sessions, and Lutfi Arikan for technical assistance.
REFERENCES 1. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Geneva, Switzerland: World Health Organization; 2000. 2. Cole TJ, Lobstein T. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity. Pediatr Obes. 2012;7:284-294. 3. Webb TL, Joseph J, Yardley L, Michie S. Using the internet to promote health behavior change: a systematic review and meta-analysis of the impact of theoretical basis, use of behavior change techniques, and mode of delivery on efficacy. J Med Internet Res. 2010;12:e4. 4. Nguyen B, Kornman KP, Baur LA. A review of electronic interventions for prevention and treatment of overweight and obesity in young people. Obes Rev. 2011;12:e298-e314. Bodurtha J, 5. Smith AJ, Skow A, Kinra S. Health information technology in screening and treatment of child obesity: a systematic review. Pediatrics. 2013;131:e894-e902. 6. Shapiro JR, Bauer S, Hamer RM, Kordy H, Ward D, Bulik CM. Use of text messaging for monitoring sugarsweetened beverages, physical activity, and screen time in children: a pilot study. J Nutr Educ Behav. 2008;40:385-391.
Journal of Nutrition Education and Behavior Volume 47, Number 1, 2015 7. Ferrer-Roca O, Cardenas A, DiazCardama A, Pulido P. Mobile phone text messaging in the management of diabetes. J Telemed Telecare. 2004;10: 282-285. 8. Anhj J, Mldrup C. Feasibility of collecting diary data from asthma patients through mobile phones and SMS (short message service): response rate analysis and focus group evaluation from a pilot study. J Med Internet Res. 2004;6:e42. 9. Rodgers A, Corbett T, Bramley D, et al. Do u smoke after txt? Results of a randomised trial of smoking cessation using mobile phone text messaging. Tob Control. 2005;14:255-261. 10. Kauer DS, Reid CS, Crooke DAH, et al. Self-monitoring using mobile phones in the early stages of adolescent depression: randomized controlled trial. J Med Internet Res. 2012;14:e67. 11. Bauer S, Okon E, Meermann R, Kordy H. Technology-enhanced maintenance of treatment gains in eating disorders: efficacy of an intervention delivered via text messaging. J Consult Clin Psychol. 2012;80:700-706. 12. de Niet J, Timman R, Bauer S, et al. The effect of a short message service maintenance treatment on body mass index and psychological well-being in overweight and obese children: a randomized controlled trial. Pediatr Obes. 2012;7:205-219. 13. Bauer S, de Niet J, Timman R, Kordy H. Enhancement of care through self-monitoring and tailored feedback via text messaging and their use in the treatment of childhood overweight. Patient Educ Couns. 2010;79:315-319. 14. Cardoso G, Gomes MdC, Espanha R, Lapa T. E-Generation: Os Usos de Media Pelas Crianc¸as e Jovens em Portugal. [E-Generation: The Use of Media by Children and Adolescents in Portugal]. Lisbon, Portugal: CIES/ISCTE; 2007. 15. Bauer S, Percevic R, Okon E, Meermann R, Kordy H. Use of text messaging in the aftercare of patients with bulimia nervosa. Eur Eat Disord Rev. 2003;11:279-290. 16. Bandura A. Social cognitive theory. In: Vasta R, ed. Annals of Child Development. Greenwich, CT: JAI Press; 1989:1-60. 17. Currie C, Gabhainn SN, Godeau E, et al. Inequalities in children’s health: HBSC,
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
International Report from the 2005/2006 Survey. Geneva, Switzerland: World Health Organization; 2008:83-109. Vereecken CA, Maes LA. Belgian study on the reliability and relative validity of the Health Behaviour in School-Aged Children food-frequency questionnaire. Public Health Nutr. 2003;6: 581-588. Golan M, Weizman A. Reliability and validity of the Family Eating and Activity Habits Questionnaire. Eur J Clin Nutr. 1998;52:771-777. Cole TJ. A chart to link child centiles of body mass index, weight and height. Eur J Clin Nutr. 2002;56:1194-1199. Tudor-Locke C, Pangrazi RP, Corbin CB, et al. BMI-referenced standards for recommended pedometer-determined steps/day in children. Prev Med. 2004;38: 857-864. Tudor-Locke C, Hatano Y, Pangrazi RP, Kang M. Revisiting ‘‘how many steps are enough?’’. Med Sci Sports Exerc. 2008;40(7 suppl): S537-S543. World Health Organization. Fruit and Vegetables Promotion Initiative. Geneva, Switzerland: World Health Organization; 2003. Conover WJ. Practical Nonparametric Statistics. 3rd ed. New York, NY: Wiley; 1999. Wei J, Hollin I, Kachnowski S. A review of the use of mobile phone text messaging in clinical and healthy behaviour interventions. J Telemed Telecare. 2011;17:41-48. Epstein LH, Paluch RA, Kilanowski CK, Raynor HA. The effect of reinforcement or stimulus control to reduce sedentary behavior in the treatment of pediatric obesity. Health Psychol. 2004;23:371-380. Williams SL, French DP. What are the most effective intervention techniques for changing physical activity selfefficacy and physical activity behaviour—and are they the same? Health Educ Res; 2011:308-322. Lau PWC, Lau EY, Wong DP, Ransdell L. A systematic review of information and communication technology-based interventions for promoting physical activity behavior change in children and adolescents. J Med Internet Res. 2011;13:e48.