Sage Journals: Discover world-class research

Abstract

Background

Interventions using commercial digital health tools do favorably affect health outcomes. However, the effect of digital tools on cardiorespiratory fitness, a more novel indicator cardiovascular risk, is unclear.

Purpose

Synthesize the digital health intervention literature and answer the following question: What is the effect of interventions using mobile health apps, wearable activity trackers, and/or text messaging on cardiorespiratory fitness?

Methods

A systematic review and a meta-analysis (PROSPERO CRD42023423925) were conducted to evaluate the immediate digital health intervention effect on adult cardiorespiratory fitness. In March 2023, a search of databases Embase, MEDLINE, CINHAL, and Cochrane Library was completed. Studies were included if the intervention used a mobile health app, text messaging, and/or activity tracker. Studies were excluded if an objective measure of fitness was not used; the sample included children; the setting was hospital-based; and the digital health technology was only used for data collection or described as virtual reality. Using a random-effects model, two separate meta-analyses were completed: one for single-group studies and one for multi-group studies. Standardized mean difference effect sizes (Cohen's d) were calculated. Study quality was evaluated with the Cochrane Risk of Bias tool and ROBINS-I tool.

Results

Fifty-three studies (3657 individuals) with pre-post designs (12 single-group, 41 multi-group) were included. Most studies targeted participants with a specific chronic health condition. Digital health interventions in the single-group studies had a moderate-to-large effect size (d = 0.62, 95% confidence interval (CI) [0.41–0.84], p < 0.001), and multi-group studies had small-to-moderate effect size (d = 0.38, 95% CI 0.21–0.55, p < 0.001). Significant heterogeneity of effects was observed in both the single-group and multi-group studies.

Conclusions

Interventions using text messaging, a mobile app, or activity tracker alone or in combination are effective in improving cardiorespiratory fitness in adults, particularly for those with a chronic health condition.

Keywords

Cardiorespiratory fitness digital health digital tools meta-analysis physical activity

Introduction

Digital health tools, including wearable physical activity monitors, smartphone health apps, and text messaging, are effective intervention components to improve health behaviors and health outcomes.^1–7 For example, meta-analytic evidence indicates that weight loss intervention participants lost seven times more weight when text messaging was included as an intervention component.⁸ Likewise, among participants with type II diabetes, brief text messaging interventions had a significant favorable effect on hemoglobin A1C (3-month average blood glucose) compared to usual care conditions.⁹ Other effective digital health strategies include mobile app use for informational support and as a tool to self-monitor weight and/or physical activity behavior.^1,10 Digital health interventions commonly evaluate physical activity behaviors (i.e., steps, duration, intensity) and cardiovascular outcomes, including body mass index, blood pressure, cholesterol, and glycosylated hemoglobin (A1C). Cardiorespiratory fitness, a powerful predictor of cardiovascular risk, is a relatively novel outcome of interest in health behavior research. High-level evidence of the effect of digital health tools on fitness in adults is limited. Emberson et al. (2021) conducted a systematic review evaluating the effects of smartphone-based physical activity interventions; only one of the 20 studies included in the review evaluated fitness as an outcome.¹¹ In that one randomized design study with low-activity adults in New Zealand (N = 51), there was no significant difference in fitness evaluated with a 1 mile walk/run test compared to control.¹¹ In another recent systematic review with meta-analysis, the effect of smartphone-based interventions on fitness in adults with coronary artery disease included eight studies (1120 participants).¹² Here, where the fitness measure (6 min walk test) was the same across all eight studies reviewed, fitness was significantly better among those receiving the smartphone app intervention versus those who received standard cardiac rehab.

Cardiorespiratory fitness is the efficiency of the heart and lungs working together to supply oxygen to muscles in the body during exertional physical activity. Common measures of fitness include volume of oxygen consumed during activity (VO₂) and metabolic equivalents (METs), which are the metabolic cost of an activity compared to resting metabolism. In adults, optimal health is associated with performing activities at a MET level of 8 to 10; increased risk of cardiovascular disease and all-cause mortality are associated with only performing activities with a MET level less than 5.¹³ The gold standard method of fitness evaluation is a graded exercise test (on treadmill or bicycle) that progresses until maximum effort or exhaustion is achieved while the individual wears a mask to measure oxygen and carbon dioxide exchange.¹⁴ Reliable and valid submaximal fitness evaluation, which is more amenable to clinical and community settings, has been developed and include walk, step, and cycle tests.^15,16 As a biomarker, submaximal fitness tests are sensitive to intervention and therefore useful for serial evaluation of efforts to change physical activity behavior.¹⁷ Since 2016, the American Heart Association has called for fitness assessment in adults at least annually during routine clinical visits to improve cardiovascular risk prediction and risk management by objectively evaluating the cardiorespiratory efficiency.¹³ Given the surge in digital health research and the increasing focus on cardiorespiratory fitness, description of the effect of digital health interventions on fitness outcomes is needed to support evidence-based practice. The purpose of this systematic review and meta-analysis was to synthesize the current digital health intervention literature and answer the following question: what is the effect of interventions using mobile health apps, wearable activity trackers, and/or text messaging on cardiorespiratory fitness?

Methods

The PRISMA guidelines¹⁸ guided the protocol development for this systematic review and meta-analysis, and the protocol is registered with PROSPERO (CRD42023423925).

Information sources

Search strategy

The literature was searched using search strategies developed by a health sciences librarian (DW) on the topics of mobile or digital health behavior interventions, physical activity, and cardiorespiratory fitness in adults. These strategies were developed using a combination of standardized index terms and keywords and were performed in Embase (Elsevier), MEDLINE (EBSCO), CINHAL (EBSCO), and Cochrane Library (Wiley). Search limits applied were published in English, published between January 2000 and December 2023, adult sample, and all study types, except for reviews, letters, notes, and editorials. The publication timeframe was set according to two key digital health milestones; mobile phone data interaction functionality (email and text) began in 2000, and Fitbit, the first commercial wearable activity tracker, became available in 2009.¹⁹ The databases were searched, and articles were retrieved on 31 March 2023. The search strategies used are available to view in Supplemental file 1.

Eligibility criteria

Studies were included if they reported on interventions that used a mobile health app, text messaging, and/or a wearable activity tracker and an objective measure of cardiorespiratory fitness. Studies were excluded if an objective measure of fitness was not used; the sample included children; the setting was hospital based; and the digital health technology was only used for data collection or was described as virtual reality.

Selection process

Using Covidence software,²⁰ two doctorally prepared researchers (SR and CB) independently reviewed titles and abstracts to confirm eligibility, then together, the researchers reconciled all discrepancies. Figure 1 details the study selection process.

Figure 1.

PRISMA flowchart.

Data collection

Coding

An a priori codebook was developed from prior physical activity studies performed by the first author, including a meta-analysis on interventions targeting sedentary behavior in older adults.^21–23 To evaluate the codebook performance with data extraction, the codebook was piloted independently by SR and CB with 10 studies, and then revised following consensus. All eligible studies were independently coded by SR and CB using Covidence software, and all discrepancies reconciled through discussion and re-review of the associated study.

Data items

Descriptive data on each study (Supplemental Table 2) were collected, including study location, setting, funding status, study design, sample demographics (i.e., age, gender, ethnicity), and intervention characteristics (i.e., duration, digital health components, behavior change components). Select taxonomy-defined behavior change components collected included self-monitoring, goal-setting, feedback, problem solving, and role modeling,²⁴ as well as shared and supervised physical activities. Data needed to calculate effect sizes were also collected, including sample size, outcome mean (most proximal to end of intervention), and measures of variability. If a study did not report data needed to calculate effect sizes, a request for the missing data was made to the corresponding author. A total of 21 corresponding authors were contacted for missing data; 28.5% (n = 6) responded with the data needed to include their study in our meta-analysis.

Bias assessment

The Cochrane Risk of Bias tool was used to evaluate the quality of studies using a randomized control trial (RCT) design.²⁵ The Cochrane tool evaluates bias as high, low, or unclear across five domains, namely, selection, allocation concealment, reporting, and other sources. The ROBINS-I tool was used to evaluate the quality of non-randomized design studies.²⁶ The ROBINS-I tool evaluates bias as low, moderate, critical, or no information across seven domains, including confounding, participant selection, intervention classification, intervention deviation, missing data, and measurements. Both tools have an overall bias rating. Researchers (SR and CB) evaluated the quality of each study independently, then discrepancies were reconciled through discussion. Publication bias was also evaluated with the trim and fill test,²⁷ Egger's regression intercept,²⁸ and funnel plots²⁹ using Comprehensive Meta-Analysis software (CMA). Multiple publication bias tests are done in order to provide greater confidence in the results.³⁰

Data analysis

Data analysis was accomplished with CMA software.³¹ A random effects-model was chosen due to the expected heterogeneity among the studies.³⁰ Two separate meta-analyses were completed: one for single-group studies and one for multi-group studies using either randomized or non-randomized designs.³⁰ Only objective measures of cardiorespiratory fitness were used. Standardized mean difference effect sizes were calculated in CMA, with the majority done by subtracting the mean of the control group from the mean of the treatment group and dividing by the pooled standard deviation. In the multi-group meta-analysis, two studies required the use of difference in means and p values;^32,33 one study required sample size and p value;³⁴ and one study required means and t value.³⁵ In the single-group meta-analysis, the study involved three intervention groups who all used pedometers and were combined into one group. As per our inclusion criteria, every group had a wearable device.³⁶ Where studies had more than one outcome measure, they were combined in CMA by taking the average of the means.³⁰ The type of fitness outcome measure can be seen in the forest plots for each meta-analysis (Figures 2 and 3), including subjective, objective, or combined.

Figure 2.

Forest plot of single-group, posttest cardiorespiratory fitness outcomes among adults. A random-effects model was used to calculate effect sizes. Square areas are proportional to the study weight.

Figure 3.

Forest plot of multi-group, posttest cardiorespiratory fitness outcomes among adults. A random-effects model was used to calculate effect sizes. Square areas are proportional to the study weight.

Cohen's d was chosen to quantify the effect sizes as recent research has questioned the bias of Hedge's g.³⁷ A positive effect size was set to indicate improvement in cardiorespiratory outcomes in the intervention group compared to the control group or to themselves pre-intervention. A negative effect size indicates worse cardiorespiratory outcomes in the intervention group compared to the controls or compared to themselves pre-intervention. Effect sizes are classified as small (≤0.20), medium (=0.50), or large (≥0.80).³⁸ Heterogeneity of effects was assessed using prediction intervals (PI), T, and I². Q is no longer recommended as an accurate measure of heterogeneity.³⁹ A one-study-removed sensitivity analysis was also done for each meta-analysis to identify outliers.

Results

Study characteristics

A total of 53 distinct studies published between 2014 and 2023 were included in the narrative synthesis. All studies collected pre-post measures; 12 used a single-group design, and 41 used a multi-group design. Of those testing more than one group, 85% (n = 35) randomized participants to groups. Studies were conducted in North America (n = 23), Europe (n = 15), Asia (n = 8), Australia (n = 6), Africa (n = 1), and the Middle East (n = 1).

Sample characteristics

The single-group studies (12 studies with a total of 699 individuals) were more diverse with race compared to the multi-group studies (41 studies with a total of 2958 individuals); participants were 47.4% non-White in the single-group studies and 10.4% non-White in the multi-group studies. The mean age of participants in the single-group studies was 48.0 (SD 7.4) years and 49.5 (SD 15.3) years in the multi-group studies. Females comprised 64.1% of the sample in the single-group studies and 55.5% of the samples in the multi-group studies. Most of the studies (n = 44; 78.5%) targeted participants with a specific chronic health condition or problem, including cancer, diabetes, obesity, cardiovascular disease, liver disease, kidney disease, substance use disorder, or psychological disorder. The sample size range was 10–288 in the single-group studies and 11–335 in the multi-group studies.

Intervention characteristics

Across studies, intervention duration was between 4 and 168 weeks (3.2 years), with most delivering a 12-week intervention. Most studies delivered the intervention in the home setting (n = 42; 76%) or a combination of home and elsewhere (i.e., clinic, community center). Interventions were also delivered in clinical settings (n = 3),^40–42 workplaces (n = 3),^43–45 community settings (n = 9),^{33,36,46–51} and, all or in part, at an academic setting (n = 3).^52–54 Most studies used a combination of digital health intervention strategies; four studies used text messaging and an activity tracker;^32,44,53,55 four studies used text messaging and an app;^23,46,56,57 14 studies used an app and a wearable activity tracker;^{42,47,48,52,58–68} and nine studies used text messaging, an app, and activity tracker.^{35,51,69–75} Only two studies used text messaging alone.^45,76 A number of studies used only an app^{33,46,48–50,54,68,77–82} or only an activity tracker.^36,83–90 Supplemental Table 2 details the behavior change strategies used along with the digital health tools, including goal setting, feedback, barrier problem solving, shared physical activity, supervised physical activity, and motivational interviewing.

Study quality of the 35 randomized design studies were evaluated with the Cochrane Risk of Bias Tool. While there were no studies evaluated as “high risk” on overall risk of bias, 45.7% (n = 16) were evaluated as “low risk,” and 54.2% (n = 19) did not report enough information to render an overall bias judgement. Evaluation of the 18 non-randomized studies using the ROBINS-I risk of bias tool found a range related to the overall risk of bias, including low risk (n = 2), moderate risk (n = 2), serious risk (n = 6), and not enough information (n = 8).

Outcome measures

All included studies used an objective measure of cardiorespiratory fitness (inclusion criteria); however, a total of 19 studies (35%) also used a subjective measure of physical activity behavior. The International Physical Activity Questionnaire (IPAQ), a measure of physical activity level, was the most common subjective measure of physical activity behavior used. Objective measures used to evaluate cardiorespiratory fitness included maximal exertion walking or cycling tests and submaximal exertion walking or stepping tests.

Effect of digital health interventions on cardiorespiratory fitness

A summary of effect sizes for both the single group and multi-group meta-analyses is presented in Figures 2 and 3, respectively. Digital health interventions in the single-group studies (k = 12, n = 12) had a statistically significant, moderate-to-large effect size (d = 0.60, 95% CI [0.39–0.81], p < 0.001) on cardiorespiratory fitness (Figure 2). Effect sizes ranged from 0 to 3.40, with no negative effect sizes. As expected, there was a statistically significant heterogeneity of effects across the studies 0.09–1.28, I² = 73.77%, T = 0.29, p < 0.001). Digital health interventions in the multi-group studies (k = 41, n = 41) had a small-to-moderate effect size (d = 0.35, 95% CI 0.17–0.53, p < 0.001) (Figure 3). Effect sizes ranged from −0.36 to 2.05, with a positive effect size indicative of the intervention group having better cardiorespiratory fitness outcomes compared to the control group. Heterogeneity was again at statistically significant levels, as expected (95% PI −0.67–1.37, I²= 78.72%, T = 0.50, p < 0.001).

The one-study-removed sensitivity analysis demonstrated that no studies had a significant influence on the magnitude of the summary effect sizes in either meta-analysis. Despite sensitivity analysis not identifying any outliers, we also completed the single-group meta-analysis without Nabutovsky et al. (2020), the largest effect size in the study. This reduced the summary effect size by approximately to 0.52, a minimal change. Pre-post correlations of 0.5 were used for all studies after sensitivity analysis was done by setting correlations to 0.01, and then 0.09 with no change observed in the effect sizes.

The funnel plots for the single-group meta-analysis (Figure 4) and multi-group meta-analysis (Figure 5) are generally symmetrical, though we note few studies in the lower left corner of the plots representing studies with smaller effect sizes, which would hypothetically demonstrate less publication bias. Egger's regression test in the single-group meta-analysis was not significant (t = 1.16, p = 0.27, intercept 1.21), but was significant in the multi-group meta-analysis (t = 2.01, p = 0.05, intercept 1.49). This suggests that the single-group data set was unlikely to be influenced by small sample bias, but the multi-group data may have been.⁹¹ Both of the trim and fill tests returned no studies trimmed, and, therefore, the point estimates and funnel plots remained exactly the same, indicating less likelihood of publication bias.²⁷

Figure 4.

Forest plot of single-group, posttest cardiorespiratory fitness outcomes among adults. A random-effects model was used to calculate effect sizes. Square areas are proportional to the study weight.

Figure 5.

Forest plot of multi-group, posttest cardiorespiratory fitness outcomes among adults. A random-effects model was used to calculate effect sizes. Square areas are proportional to the study weight.

Moderator analyses

Moderator analyses were completed to investigate heterogeneity (Table 1). Subgroup analyses included United States versus international location, sample mean age greater or less than 45.0, randomized design, if the study included individuals with chronic conditions or not, and multiple intervention characteristics, including inclusion of physical activity education, self-tracking of physical activity, encouragement, goal-setting, feedback on progress or technique, addressing barriers to physical activity, peer role modeling, shared peer physical activity, supervised physical activity, motivational interviewing, text messaging, wearable device, and use of an app. Meta-regression was done to determine if the duration of the intervention or age of participants affected its efficacy.

Table 1.

Moderator analyses.

Subgroup analysis
Single-group Studies	Cohen's d	k	p value	95% CI	Q_between	p of Q
Mean age ≤45.0 versus	0.48***	4	<0.001	0.27–0.69	1.10	0.29
>45.0 years old	0.68***	8	<0.001	0.41–1.01
Barriers to PA addressed versus	1.39**	4	0.002	0.52–2.61	10.26**	<0.01
not addressed	0.43***	7	<0.001	0.33–0.53
Shared peer activity versus	1.12**	5	0.004	0.35–1.89	2.47	0.12
None	0.50***	7	<0.001	0.37–0.62
Use of app versus	0.72**	8	0.001	0.28–1.16	0.57	0.45
no app	0.58***	4	<0.001	0.39–0.70
US-based versus	0.59**	8	0.001	0.24–0.94	0.08	0.78
international	0.65***	4	<0.001	0.42–0.88
Multi-group studies
US versus	0.05	11	0.78	−0.28–0.37	4.43*	0.04
international	0.44***	30	<0.001	0.26–0.62
Randomized versus	0.34**	34	0.002	0.13–0.55	0.25	0.62
non-randomized	0.38***	7	<0.001	0.23–0.53
Mean age ≤45.0 versus	0.43	24	0.06	−0.01–0.87	1.90	0.39
>45.0 years old	0.35***	30	<0.001	0.21–0.50
Chronic conditions versus	0.30**	33	0.001	0.12–0.47	53.21***	<0.001
no chronic conditions	0.27	7	0.09	−0.04–0.57
Goal setting versus	0.32*	33	0.003	0.11–0.54	0.82	0.37
none	0.46***	8	<0.001	0.25–0.67
Feedback versus	0.25*	33	0.004	0.08–0.41	4.40	0.11
none	0.91*	7	0.004	0.29–1.54
Barriers to PA addressed	0.25	10	0.19	−0.12–0.61	1.61	0.45
versus not addressed	0.40***	30	<0.001	0.18–0.62
Shared peer activity versus	0.38***	6	<0.001	0.17–0.59	0.04	0.84
none	0.38**	35	0.001	0.14–0.56
Supervised PA versus	0.39***	10	<0.001	0.23–0.56	0.13	0.72
unsupervised	0.38*	31	0.005	0.10–0.57
Text messaging versus	0.28	16	0.13	−0.06–0.46	1.96	0.16
none	0.44***	25	<0.001	0.23–0.67
Wearable device versus	0.31*	26	0.008	0.08–0.55	0.24	0.63
none	0.41*	15	0.007	0.11–0.70
Use of app versus	0.36**	30	0.001	0.16–0.56	0.01	0.92
no app	0.33	11	0.09	−0.05–0.71

Meta-regression
Single-group studies	Coeff.	n	p value	95% CI
Duration	0.56	9	0.84	−0.02–0.03
Mean age	−0.43	9	0.40	−0.03–0.08
Multi-group studies
Duration	0.41	19	0.99	−0.04–0.86
Mean age	0.40	19	0.98	−0.02–0.02

PA: physical activity; Coeff: coefficient.

*p < 0.05, **p < 0.01, ***p < 0.001.

In the single-group studies, there was one significant moderator finding when examining subgroups. Studies that included addressing participants’ barriers to physical activity (k = 4, n = 4) had a statistically significant higher effect size (d = 1.39, 95% CI 0.52–2.26, p = 0.002) compared to studies, which did not (d = 0.43, k = 7, n = 7, 95% CI 0.33–0.53, p < 0.001, Q = 10.26, p of Q < 0.01). There were no statistically significant differences between interventions with shared physical activity, those inclusive of a mobile app versus not, or supervised versus unsupervised peer activity, location of study, and there were no differences based on the average participant age (Table 1). Participants with chronic conditions, peer modeling, education, self-tracking of activity, encouragement, feedback, goal setting, motivational interviewing, text messaging, and use of a wearable device (if reported) did not have enough studies in each subgroup (minimum k = 4) to perform a reliable analysis.

In the multi-group studies, there were two significant moderator findings. Studies that included individuals with chronic conditions (e.g. diabetes) had a significantly higher effect size (d = 0.30, k = 33, n = 33, 95% CI 0.12–0.47, p = 0.001) than studies, which did not (d = 0.27, k = 7, n = 7, 95% CI −0.04–0.57, p = 0.09, Q = 53.21, p of Q < 0.001). Studies conducted in the US had a significantly lower effect size (d = 0.05, k = 11, n = 11, 95% CI −0.28–0.37, p = 0.78) than those conducted elsewhere (d = 0.35, k = 30, n = 30, 95% CI 0.19–0.51, p < 0.001, Q = 4.43, p of Q = 0.04). Studies with and without individuals with chronic conditions, randomization, mean age, goal setting, feedback, addressing barriers, shared peer activity, supervised physical activity, text messaging, a wearable device, and use of an app were not statistically significant compared to their absence. Passive versus active control groups and use of education, self-tracking, peer modeling, subjective outcome alone, and motivational interviewing did not have enough studies in each subgroup (minimum k = 4) to perform reliable analyses.

In each meta-analysis, a meta-regression with the Knapp–Hartung correction was completed to determine if the duration of the intervention or mean age was correlated with effect size. Neither of the analyses had statistical significance (Table 1). With removal of duration outliers (Maltais et al. (2019) in the single-group studies and Ma et al. (2019) in the multi-group studies), the statistical significance did not change.

Discussion

This systematic review and meta-analysis advance our understanding of digital health interventions by examining their effect on cardiorespiratory fitness, a novel and increasingly recognized important health outcome. Cardiorespiratory fitness (heart–lung efficiency) is one of the most important correlates of general health status and a powerful predictor of all-cause mortality independent of age, ethnicity, body mass index, cholesterol, type II diabetes, and smoking.^13,92 The American Heart Association now recommends routine measurement of fitness to both evaluate and modify cardiovascular risk. In our study, digital health interventions using text messaging, a wearable activity tracker, and/or a smartphone app had a significant favorable effect on fitness with a moderate to large effect in single-group studies and small to moderate effect in multi-group studies. These findings are good news as even small changes in fitness are clinically significant. A 1-MET increase in fitness (corresponding to V0₂ of 3.5 ml/kg/min during exercise) is associated with a 10–30% reduction in cardiovascular mortality.¹³

Few included studies tested the effect of text messaging alone, so it remains unclear whether text messaging alone can favorably affect cardiorespiratory fitness. It is important to note, cardiorespiratory fitness did significantly improve with the use of only an app or use of only an activity tracker. While the combination of the two strategies did appear helpful, the improvement in fitness was not significant. Health behavior researchers and health care providers can be reassured that digital health interventions need not be overly complex or multi-component to have a favorable effect on cardiorespiratory fitness, particularly for those with a chronic health condition. It may be that complex, intensive, multi-component health behavior strategies are needed for primary prevention and not as needed for secondary or tertiary prevention. In other words, a chronic health condition may be a strong motivator to overcome barriers to increasing physical activity. Our moderator analysis findings among the single-group studies suggest two key non-digital components to consider as part of a digital health intervention: shared physical activity and addressing participant barriers to physical activity.

Studies included in this review had samples with a wide range of chronic health conditions, including cardiovascular disease, cancer, and severe psychological conditions, and, notably, there were only two negative effect sizes. There was, however, considerable heterogeneity of effects in both the single-group and multi-group studies. This is likely due to the wide variation in study design, interventions, and patient populations. While the study samples were diverse with respect to sex and health conditions, they were not diverse with respect to race or ethnicity, particularly when a multi-group design was used. There is a clear gap in high-level evidence (i.e., randomized controlled trials) for the use of digital health interventions to affect fitness among non-White samples. Given cardiovascular morbidity and mortality outcomes are worse for the Black and Hispanic/Latino adults, digital health research with ethnically and racially diverse samples is greatly needed.^93–95

Limitations

We were unable to ascertain the study quality for approximately half of the single-group studies and half of the multi-group studies. No studies reported on baseline outcome blinding of participants. The effect of knowing your fitness level prior to a digital health intervention could not be explored. Small sample bias was likely in the multi-group meta-analysis. Small sample sizes also prevented a comprehensive analysis of moderators of intervention effects. The lack of ethnic and racial diversity among samples limits the generalizability of the findings to populations under-represented in digital health research. This study also evaluated the most proximal intervention effect. Future research on the long-term effect of digital health interventions on cardiorespiratory fitness is needed. Nonetheless, our findings provide a rigorous comprehensive analysis of a diverse range of studies using digital health strategies to affect cardiorespiratory fitness. Most multi-group studies used some form of control group and randomized participants. The findings also appear to be robust given the minimal impact on the summary of effect sizes with removal of outlier studies.

Conclusion

Digital health interventions using text messaging, a smartphone app, or activity tracker alone or in combination are effective at improving cardiorespiratory fitness in adults, particularly those with a chronic health condition. Future research is needed to determine the moderating effects of non-digital strategies commonly used along with digital health interventions (i.e., goal setting, feedback, shared physical activity). Research evaluating the long-term effect of digital health interventions as well as the effect among non-White populations is also needed. Lastly, digital health researchers are called to report on common quality metrics, such as participant and personnel blinding.

Supplemental Material

sj-docx-1-dhj-10.1177_20552076241282381 - Supplemental material for Effectiveness of digital health interventions to increase cardiorespiratory fitness: A systematic review and meta-analysis

Supplemental material, sj-docx-1-dhj-10.1177_20552076241282381 for Effectiveness of digital health interventions to increase cardiorespiratory fitness: A systematic review and meta-analysis by Sheri Rowland, Christina Bach, Krystyna Simon and Danielle M Westmark, Edie Sperling in DIGITAL HEALTH

Supplemental Material

sj-docx-2-dhj-10.1177_20552076241282381 - Supplemental material for Effectiveness of digital health interventions to increase cardiorespiratory fitness: A systematic review and meta-analysis

Supplemental material, sj-docx-2-dhj-10.1177_20552076241282381 for Effectiveness of digital health interventions to increase cardiorespiratory fitness: A systematic review and meta-analysis by Sheri Rowland, Christina Bach, Krystyna Simon and Danielle M Westmark, Edie Sperling in DIGITAL HEALTH

Footnotes

Contributorship

Sheri Rowland: conceptualization, funding acquisition, methodology, investigation, supervision, writing original draft, and writing review and editing. Christina Bach: investigation and writing review and editing. Krystyna Simon: investigation and writing review and editing. Danielle M. Westmark: investigation and writing review and editing. Edie Sperling: methodology, investigation, formal analysis, writing original draft, and writing review and editing.

Data availability

Data are available upon request to the corresponding author.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical approval

This study is a review of published anonymized human study research. The Institutional Review Board of the University of Nebraska Medical Center waived the need for review.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by Rowland's faculty development funds at the University of Nebraska Medical Center College of Nursing.

Guarantor

SR.

ORCID iDs

Sheri Rowland

Krystyna Simon

Edie Sperling

Supplemental material

Supplemental material for this article is available online.

References

Mateo

Granado-Font

Ferré-Grau

, et al. Mobile phone apps to promote weight loss and increase physical activity: a systematic review and meta-analysis. J Med Internet Res 2015; 17: e253.

Changizi

Kaveh

. Effectiveness of the mHealth technology in improvement of healthy behaviors in an elderly population—a systematic review. mHealth 2017; 3: 51.

Han

Lee

. Effectiveness of mobile health application use to improve health behavior changes: a systematic review of randomized controlled trials. Healthc Inform Res 2018; 24: 207–226.

Brickwood

K-J

Watson

O’Brien

, et al. Consumer-based wearable activity trackers increase physical activity participation: systematic review and meta-analysis. JMIR MHealth UHealth 2019; 7: e11819.

Laranjo

Ding

Heleno

, et al. Do smartphone applications and activity trackers increase physical activity in adults? Systematic review, meta-analysis and metaregression. Br J Sports Med 2021; 55: 422–432.

Ferguson

Olds

Curtis

, et al. Effectiveness of wearable activity trackers to increase physical activity and improve health: a systematic review of systematic reviews and meta-analyses. Lancet Digit Health 2022; 4: e615–e626.

Armanasco

Miller

Fjeldsoe

, et al. Preventive health behavior change text message interventions: a meta-analysis. Am J Prev Med 2017; 52: 391–402.

Siopis

Chey

Allman-Farinelli

. A systematic review and meta-analysis of interventions for weight management using text messaging. J Hum Nutr Diet 2015; 28: 1–15.

Arambepola

Ricci-Cabello

Manikavasagam

, et al. The impact of automated brief messages promoting lifestyle changes delivered via mobile devices to people with type 2 diabetes: a systematic literature review and meta-analysis of controlled trials. J Med Internet Res 2016; 18: e86.

10.

Fanning

Mullen

McAuley

. Increasing physical activity with mobile devices: a meta-analysis. J Med Internet Res 2012; 14: e2171.

11.

Emberson

Lalande

Wang

, et al. Effectiveness of smartphone-based physical activity interventions on Individuals’ health outcomes: a systematic review. BioMed Res Int 2021; 2021: 1–13.

12.

Murphy

Meehan

Koshy

, et al. Efficacy of smartphone-based secondary preventive strategies in coronary artery disease. Clin Med Insights Cardiol 2020; 14: 1–7. 117954682092740.

13.

Ross

Blair

Arena

, et al. Importance of assessing cardiorespiratory fitness in clinical practice: a case for fitness as a clinical vital sign: a scientific statement from the American Heart Association. Circulation 2016; 134: CIR-0000000000000461.

14.

Liguori

, Medicine AC of S. ACSM’s guidelines for exercise testing and prescription. Lippincott Williams & Wilkins, https://books.google.com/books?hl=en&lr=&id=rpoMEAAAQBAJ&oi=fnd&pg=PT43&dq=ACSM%27s+Guidelines+for+Exercise+Testing+and+Prescription,+11th+Edition&ots=7LN3_O3HUl&sig=QzDOSR_6XA2yzTsgo1VScirNDjI (2020, accessed 1 February 2024).

15.

Bennett

Parfitt

Davison

, et al. Validity of submaximal step tests to estimate maximal oxygen uptake in healthy adults. Sports Med 2016; 46: 737–750.

16.

Castro-Piñero

Marin-Jimenez

Fernandez-Santos

, et al. Criterion-related validity of field-based fitness tests in adults: a systematic review. J Clin Med 2021; 10: 3743.

17.

Clarke

De Lannoy

Ross

. Comparison of measures of maximal and submaximal fitness in response to exercise. Med Sci Sports Exerc 2017; 49: 711–716.

18.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg 2021; 372: n71.

19.

Istepanian

. Mobile health (m-Health) in retrospect: the known unknowns. Int J Environ Res Public Health 2022; 19: 3747.

20.

Covidence Systematic Review Software, www.covidence.org , 2024.

21.

Chase

J-AD

Otmanowski

Rowland

, et al. A systematic review and meta-analysis of interventions to reduce sedentary behavior among older adults. Transl Behav Med 2020; 10: 1078–1085.

22.

Rowland

Berg

Kupzyk

, et al. Feasibility and effect of a peer modeling workplace physical activity intervention for women. Workplace Health Saf 2018; 66: 428–436.

23.

Rowland

Ramos

Trinidad

, et al. mHealth intervention to improve cardiometabolic health in rural hispanic adults: a pilot study. J Cardiovasc Nurs 2022; 37: 439–445.

24.

Michie

Richardson

Johnston

, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med 2013; 46: 81–95.

25.

Higgins

Altman

Gøtzsche

, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. Br Med J 2011; 343: 1–9.

26.

Sterne

Hernán

Reeves

, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. Br Med J 2016; 355: 1–7.

27.

Duval

Tweedie

. A nonparametric “trim and fill” method of accounting for publication bias in meta-analysis. J Am Stat Assoc 2000; 95: 89–98.

28.

Begg

Mazumdar

. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994; 50: 1088–1101.

29.

Cohen

. Statistical power analysis for the behavioral sciences. New York, NY, USA: Routledge, 2013.

30.

Borenstein

Hedges

Higgins

, et al. Introduction to meta-analysis. Hoboken, NJ, USA: John Wiley & Sons, 2021.

31.

Borenstein

Hedges

Higgins

, et al. Comprehensive Meta-Analysis Version 4.

32.

Jakicic

Davis

Rogers

, et al. Effect of wearable technology combined with a lifestyle intervention on long-term weight loss: the IDEA randomized clinical trial. Obstet Gynecol Surv 2017; 72: 27–28.

33.

Plotnikoff

Jansson

Duncan

, et al. mHealth to support outdoor gym resistance training: the ecofit effectiveness RCT. Am J Prev Med 2023; 64: 853–864.

34.

Christensen

Sundberg

Osterkamp

, et al. Interval walking improves glycemic control and body composition after cancer treatment: a randomized controlled trial. J Clin Endocrinol Metab 2019; 104: 3701–3712.

35.

Aschbrenner

Naslund

Gorin

, et al. Group lifestyle intervention with mobile health for young adults with serious mental illness: a randomized controlled trial. Psychiatr Serv Wash DC 2022; 73: 141–148.

36.

Wilbur

Miller

Fogg

, et al. Randomized clinical trial of the women’s lifestyle physical activity program for African-American women: 24- and 48-week outcomes. Am J Health Promot AJHP 2016; 30: 335–345.

37.

Lin

Aloe

. Evaluation of various estimators for standardized mean difference in meta-analysis. Stat Med 2021; 40: 403–426.

38.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Br Med J 2021; 372: 1–9.

39.

Borenstein

. Common mistakes in meta-analysis and how to avoid them. Biostat, https://cir.nii.ac.jp/crid/1131412883849101322, 2019.

40.

Aschbrenner

Naslund

Shevenell

, et al. A pilot study of a peer-group lifestyle intervention enhanced with mHealth technology and social media for adults with serious mental illness. J Nerv Ment Dis 2016; 204: 483–486.

41.

Berglind

Nyberg

Willmer

, et al. An eHealth program versus a standard care supervised health program and associated health outcomes in individuals with mobility disability: study protocol for a randomized controlled trial. Trials 2018; 19: 258.

42.

Browne

Battaglini

Jarskog

, et al. Targeting physical health in schizophrenia: results from the physical activity can enhance life (PACE-Life) 24-week open trial. Ment Health Phys Act 2021; 20: 1–11.

43.

Howie

Daniels

. Physical activity, sleep, fitness, and cognitive performance of university personnel: needs for exercise is medicine. 2021 ACSM Annual Meeting & World Congresses [virtual], June 1–5, 2021. Med Sci Sports Exerc 2021; 53: 460–460.

44.

Buchholz

Ingram

Wilbur

, et al. Bilingual Text4Walking food service employee intervention pilot study. JMIR MHealth UHealth 2016; 4: e68.

45.

Dunning

McVeigh

Goble

, et al. The effect of interrupting sedentary behavior on the cardiometabolic health of adults with sedentary occupations: a pilot study. J Occup Environ Med 2018; 60: 760–767.

46.

Bonato

Turrini

De Zan

, et al. A mobile application for exercise intervention in people living with HIV. Med Sci Sports Exerc 2020; 52: 425–433.

47.

Linke

Hovsepians

Schnebly

, et al. The Go-VAR (veterans active recovery): an adjunctive, exercise-based intervention for veterans recovering from substance use disorders. J Psychoactive Drugs 2019; 51: 68–77.

48.

Orleans-Pobee

Browne

Ludwig

, et al. Physical activity can enhance life (PACE-life): results from a 10-week walking intervention for individuals with schizophrenia spectrum disorders. J Ment Health 2022; 31: 357–365.

49.

Plotnikoff

Wilczynska

Cohen

, et al. Integrating smartphone technology, social support and the outdoor physical environment to improve fitness among adults at risk of, or diagnosed with, type 2 diabetes: findings from the ‘eCoFit’ randomized controlled trial. Prev Med 2017; 105: 404–411.

50.

Wilczynska

Jansson

Lubans

, et al. Physical activity intervention for rural middle-aged and older Australian adults: a pilot implementation study of the ecofit program delivered in a real-world setting. Pilot Feasibility Stud 2021; 7: 81.

51.

Muntaner-Mas

Vidal-Conti

Borràs

, et al. Effects of a Whatsapp-delivered physical activity intervention to enhance health-related physical fitness components and cardiovascular disease risk factors in older adults. J Sports Med Phys Fitness 2017; 57: 90–102.

52.

Coombes

Keating

Mielke

, et al. Personal activity intelligence (PAI) e-health program in people with type 2 diabetes: a pilot randomized controlled trial. Med Sci Sports Exerc 2022; 54: 18–27.

53.

Brooker

Gomersall

King

, et al. The efficacy of morning versus evening exercise for weight loss: a randomized controlled trial. Obesity 2023; 31: 83–95.

54.

Mora-Gonzalez

Pérez-López

Delgado-Fernández

. The “$in TIME” gamification project: using a mobile app to improve cardiorespiratory fitness levels of college students. Games Health J 2020; 9: 37–44.

55.

Van Blarigan

Dhruva

Atreya

, et al. Feasibility and acceptability of a physical activity tracker and text messages to promote physical activity during chemotherapy for colorectal cancer: pilot randomized controlled trial (Smart Pace II). JMIR Cancer 2022; 8: e31576.

56.

Varnfield

Karunanithi

Lee

C-K

, et al. Smartphone-based home care model improved use of cardiac rehabilitation in postmyocardial infarction patients: results from a randomised controlled trial. Heart 2014; 100: 1770–1779.

57.

Ochi

Tsuji

Narisawa

, et al. Cardiorespiratory fitness in breast cancer survivors: a randomised controlled trial of home-based smartphone supported high intensity interval training. BMJ Support Palliat Care 2022; 12: 33–37.

58.

Avila

Claes

Goetschalckx

, et al. Home-based rehabilitation with telemonitoring guidance for patients with coronary artery disease (short-term results of the TRiCH study): randomized controlled trial. J Med Internet Res 2018; 20: e225.

59.

Cheong

Cha

, et al. Efficacy of mobile health care application and wearable device in improvement of physical performance in colorectal cancer patients undergoing chemotherapy. Clin Colorectal Cancer 2018; 17: e353–e362.

60.

Cooes

Keating

Mielke

, et al. Personal activity intelligence e-health program in people with type 2 diabetes: a pilot randomized controlled trial. Med Sci Sports Exerc 2022; 54: 18–27.

61.

Howie

Daniels

Human

, et al. Effectiveness and implementation of a virtual versus in-person walking program among employees on physical activity, fitness, and cognition. Health Technol 2021; 11: 909–917.

62.

Keytsman

Van Noten

Spaas

, et al. Periodized home-based training: a new strategy to improve high intensity exercise therapy adherence in mildly affected patients with multiple sclerosis. Mult Scler Relat Disord 2019; 28: 91–97.

63.

Shi

, et al. Chinese home-based cardiac rehabilitation model delivered by smartphone interaction improves clinical outcomes in patients with coronary heart disease. Front Cardiovasc Med 2021; 8: 731557.

64.

Zhang

Peng

, et al. Face-to-face mentoring, remotely supervised home exercise prehabilitation to improve physical function in patients awaiting kidney transplantation: a randomized clinical trial. Front Psychol 2022; 13: 831445.

65.

McNeil

Brenner

Stone

, et al. Activity tracker to prescribe various exercise intensities in breast cancer survivors. Med Sci Sports Exerc 2019; 51: 930–940.

66.

Alibhai

SMH

Papadopoulos

Durbano

, et al. Preference-based versus randomized controlled trial in prostate cancer survivors: comparison of recruitment, adherence, attrition, and clinical outcomes. Front Oncol 2022; 12: 1033229.

67.

Park

Lee

, et al. Internet of things-based lifestyle intervention for prostate cancer patients on androgen deprivation therapy: a prospective, multicenter, randomized trial. Am J Cancer Res 2021; 11: 5496–5507.

68.

Pope

Zeng

Zhang

, et al. Effectiveness of combined smartwatch and social media intervention on breast cancer survivor health outcomes: a 10-week pilot randomized trial. J Clin Med 2018; 7: 140.

69.

Nabutovsky

Ashri

Nachshon

, et al. Feasibility, safety, and effectiveness of a mobile application in cardiac rehabilitation. Isr Med Assoc J 2020; 22: 357–363.

70.

Lahtio

Heinonen

Paajanen

, et al. The added value of remote technology in cardiac rehabilitation on physical function, anthropometrics, and quality of life: cluster randomized controlled trial. J Med Internet Res 2023; 25: e42455.

71.

Chiang

S-L

Shen

C-L

Lee

M-S

, et al. Effectiveness of a 12-week tele-exercise training program on cardiorespiratory fitness and heart rate recovery in patients with cardiometabolic multimorbidity. Worldviews Evid Based Nurs Epub ahead of print 2023;20: 339–350.

72.

Duscha

Piner

Patel

, et al. Effects of a 12-week mHealth program on peak VO(2) and physical activity patterns after completing cardiac rehabilitation: a randomized controlled trial. Am Heart J 2018; 199: 105–114.

73.

Joseph

Todd

Ainsworth

, et al. Smart walk: a culturally tailored smartphone-delivered physical activity intervention for cardiometabolic risk reduction among African American women. Int J Environ Res Public Health 2023; 20: 1000.

74.

Tulasiram

Chandrasekaran

. Are smartphones better in guiding physical activity among sedentary young adults? A randomised controlled trial. Muscles Ligaments Tendons J MLTJ 2023; 20: 1–25.

75.

Child

Kelly

Sizelove

, et al. A remote monitoring-enabled home exercise prescription for patients with interstitial lung disease at risk for exercise-induced desaturation. Respir Med 2023; 218: 107397.

76.

Pandey

Krumme

Patel

, et al. The impact of text messaging on medication adherence and exercise among postmyocardial infarction patients: randomized controlled pilot trial. JMIR MHealth UHealth 2017; 5: e110.

77.

Berglind

Yacaman-Mendez

Lavebratt

, et al. The effect of smartphone apps versus supervised exercise on physical activity, cardiorespiratory fitness, and body composition among individuals with mild-to-moderate mobility disability: randomized controlled trial. JMIR MHealth UHealth 2020; 8: e14615.

78.

Chang

S-H

Wang

L-T

Chueh

T-Y

, et al. Effectiveness of Facebook remote live-streaming-guided exercise for improving the functional fitness of community-dwelling older adults. Front Med 2021; 8: e14615.

79.

Thorsen

Yang

Valentiner

, et al. The effects of a lifestyle intervention supported by the InterWalk smartphone app on increasing physical activity among persons with type 2 diabetes: parallel-group, randomized trial. JMIR MHealth UHealth 2022; 10: e30602.

80.

Stork

Bell

Jung

. Examining the impact of a mobile health app on functional movement and physical fitness: pilot pragmatic randomized controlled trial. JMIR MHealth UHealth 2021; 9: e24076.

81.

Wang

H-L

McMillan

Vijayakumar

, et al. A behavioral physical activity intervention to manage moderate and severe fatigue among head and neck cancer patients—pre-efficacy study in the National Institutes of Health ORBIT model. Cancer Nurs 2019; 42: E1–E14.

82.

Höchsmann

Müller

Ambühl

, et al. Novel smartphone game improves physical activity behavior in type 2 diabetes. Am J Prev Med 2019; 57: 41–50.

83.

Chen

Ferrando

White

, et al. Home-based physical activity and diet intervention to improve physical function in advanced liver disease: a randomized pilot trial. Dig Dis Sci 2020; 65: 3350–3359.

84.

Gehring

Kloek

Aaronson

, et al. Feasibility of a home-based exercise intervention with remote guidance for patients with stable grade II and III gliomas: a pilot randomized controlled trial. Clin Rehabil 2018; 32: 352–366.

85.

Gordon

Bloxham

. Influence of the Fitbit charge HR on physical activity, aerobic fitness and disability in non-specific back pain participants. J Sports Med Phys Fitness 2017; 57: 1669–1675.

86.

Heron

Kee

Mant

, et al. Stroke prevention rehabilitation intervention trial of exercise (SPRITE) — a randomised feasibility study. BMC Cardiovasc Disord 2017; 17: 290.

87.

Maltais

Lemieux

AlmÉRas

, et al. One-year lifestyle intervention, muscle lipids, and cardiometabolic risk. Med Sci Sports Exerc 2019; 51: 2156–2165.

88.

Browne

Battaglini

Jarskog

, et al. Virtual group-based walking intervention for persons with schizophrenia: a pilot randomized controlled trial. Ment Health Phys Act 2023; 24: 100515.

89.

Pfirrmann

Huber

Schattenberg

, et al. Web-based exercise as an effective complementary treatment for patients with nonalcoholic fatty liver disease: intervention study. J Med Internet Res 2019; 21: e11250.

90.

Wood

Weaver

Smith-Ryan

, et al. Lessons learned from a pilot randomized clinical trial of home-based exercise prescription before allogeneic hematopoietic cell transplantation. Support Care Cancer 2020; 28: 5291–5298.

91.

Egger

Smith

Schneider

, et al. Bias in meta-analysis detected by a simple, graphical test. Br Med J 1997; 315: 629–634.

92.

Barry

Baruth

Beets

, et al. Fitness vs. fatness on all-cause mortality: a meta-analysis. Prog Cardiovasc Dis 2014; 56: 382–390.

93.

CDC. ¡A la Buena Salud! — To Good Health! Centers for Disease Control and Prevention, https://www.cdc.gov/vitalsigns/hispanic-health/index.html (2015, accessed 5 August 2018).

94.

Mehta

Armstrong

Swett

, et al. Burden of systolic and diastolic left ventricular dysfunction among Hispanics in the United States: insights from the echocardiographic study of Latinos (ECHO-SOL). Circ Heart Fail 2016; 9: 1–4.

95.

About NCHS — NCHS Fact Sheets — NCHS Data on Racial and Ethnic Disparities, https://www.cdc.gov/nchs/about/factsheets/factsheet_disparities.htm (2020, accessed 26 May 2020).

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.05 MB