A three group randomised controlled trial on effectiveness of mobile application based performance feedback interventions on cardiovascular risk factors among Nigerian university staff

Abstract

This three-group RCT investigated the comparative effectiveness of mobile application-based performance feedback (MABPF) interventions, with and without hourly reminders, on cardiovascular risk factors (CVRF) among university staff in Nigeria in the context of Mobile health technology offering promising opportunities to promote physical activity (PA), but evidence on its effectiveness in reducing CVRF, which are causes of major non-communicable diseases (NCDs), is limited. Participants were randomised to: (1) MABPF with hourly reminders, (2) MABPF without reminders, or (3) education-only control. The primary outcome was objectively measured step counts; secondary outcomes included blood pressure, adiposity indices, lipid profile, and blood sugar. The Google Fit application is MABPF, with a daily physical activity target set at 10,000 steps and 150 heart points. Pedometer step counts showed significant between-group differences, with Intervention Group 1 recording a higher median weekly step count compared to Intervention Group 2 (U = 721.500, p = 0.002) and the control group (U = 482, p < 0.001) at 12 weeks. The MABPF with reminders group showed greater improvements in systolic BP (F = 5.818, p = 0.004) and total cholesterol (F = 4.976, p = 0.008), while MABPF without reminders was more effective for blood sugar reduction (F = 3.308, p = 0.038) and HDL increase (F = 3.357, p = 0.039). Although within the control, statistically significant reductions in stress levels, increases in HDL cholesterol, and decreases in blood glucose levels were noted; however, compared to the control group, both MABPF intervention groups demonstrated superior improvements in the above stated variable as well as systolic blood pressure, total cholesterol, LDL cholesterol, and step counts. These results supports the potential of low-cost digital technology for workplace cardiovascular prevention in resource-limited settings. These findings contribute to One Health NCDs prevention by the effective CVRF reduction reported in this study using a low-cost digital technology. Furthermore, the intervention could work in other workplace settings and communities worldwide.

Trial registration: Trial number PACTR202405808298188 (24 May 2024).

Introduction

The growing burden of cardiovascular disease (CVD) represents a significant public health challenge globally, with disproportionate impacts on low- and middle-income countries, including Nigeria. The epidemiological transition underway in sub-Saharan Africa has led to a rise in non-communicable diseases (NCDs), creating unique healthcare challenges. CVDs, once considered rare in African populations, now account for a substantial proportion of morbidity and mortality across the continent, with prevalence rising particularly among urban, educated populations [1].

University staff represent a distinct occupational group with particular vulnerability to cardiovascular risk factors (CVRF) due to their predominantly sedentary work patterns, high job stress, and work environments that often discourage physical activity [2]. Research within Nigerian university settings has identified concerning rates of hypertension, dyslipidaemia, and metabolic syndrome among academic and non-academic staff, highlighting the urgent need for effective, contextually appropriate interventions [3].

Physical inactivity stands as a primary modifiable risk factor for CVD, with strong evidence supporting the role of regular physical activity (PA) in preventing and managing hypertension, dyslipidaemia, obesity, and type 2 diabetes [4]. Current guidelines recommend a minimum of 150 min of moderate-intensity aerobic activity weekly for adults, and adherence to this threshold is widely suboptimal in workplace settings where structural and cultural factors often discourage movement [5]. Fewer than 25% of working adults in sub-Saharan Africa consistently meet this benchmark [6]. Traditional approaches to PA promotion, such as structured exercise programmes and educational sessions, have shown limited effectiveness in sustaining long-term behaviour change, particularly among working adults with competing time demands [7]. The emergence of mobile health (mHealth) technologies offers novel opportunities to address physical inactivity through personalised, accessible interventions that can be integrated into daily life with minimal disruption to work responsibilities [8].

Smartphone applications that track and provide feedback on PA have demonstrated promise in promoting behaviour change across various contexts. The immediate, personalised feedback provided by such applications engages several behaviour change mechanisms, including self-monitoring, goal-setting, and reinforcement [8]. The Google Fit application, in particular, has demonstrated reliability in tracking step counts and heart points, offering users tangible metrics to monitor their PA patterns [9]. Although mobile PA monitoring applications are now well established, their integration with hourly behavioural prompts remains understudied in African workplaces facing dual disease burdens. The frequency and timing of reminders may significantly influence adherence and outcomes; hourly prompts encouraging movement throughout the workday may be especially valuable where prolonged sitting is common [10]. The effectiveness of such reminder systems has not, however, been thoroughly investigated in African workplace contexts, where environmental and cultural factors may influence intervention acceptability.

Stress represents another significant modifiable risk factor for CVD in this population. University staff commonly experience high levels of work-related stress due to academic pressures, administrative burdens, and institutional expectations [11], and PA has well-established stress-reducing effects, suggesting that mobile application-based performance feedback (MABPF) interventions may simultaneously address multiple CVRF. The integration of mHealth technologies into cardiovascular risk reduction strategies aligns with Nigeria’s National Strategic Health Development Plan, which emphasises innovation and technology in addressing NCDs [12]. Theoretical frameworks, including the Health Belief Model, Social Cognitive Theory, and the Transtheoretical Model, support the use of mobile applications for PA promotion by emphasising perceived susceptibility, self-efficacy, and staged approaches to behaviour change; all elements that can be incorporated into MABPF interventions [13]. Regular reminders may further enhance effectiveness by addressing barriers related to forgetfulness and competing priorities.

Evidence from a systematic review demonstrates that smartphone interventions increased PA by an average of 1,850 steps per day across diverse populations [14], though significant heterogeneity in implementation approaches underscores the need for context-specific research. In Nigerian settings, preliminary studies have demonstrated the feasibility and acceptability of mHealth interventions, but evidence on their effectiveness in reducing CVRF remains scarce [15]. The present study addresses these knowledge gaps by examining the effectiveness of MABPF, with and without hourly reminders, on CVRF among university staff in Nigeria, comparing these approaches with an education-only control condition, and aiming to identify the most effective strategy for reducing cardiovascular risk in this population.

This study employs the One Health framework as an analytical tool to examine how individual health outcomes interact with institutional and physical work environments. Although traditionally applied to infectious and zoonotic diseases, One Health is increasingly recognised for its relevance to chronic disease prevention, where occupational conditions, community infrastructure, and resource availability jointly shape disease burden [16, 17]. Three dimensions are applied throughout: human health, encompassing individual cardiovascular and metabolic outcomes [3]; institutional and community health, focusing on the university as a social setting whose health culture influences staff, students, and wider networks [18, 19]; and environmental context, referring to the occupational and infrastructural conditions of the Nigerian university workplace that shape CVRF prevalence and the feasibility of delivering interventions [2, 6].

Research methodology

Research design and population

This study employs a pre-test post-test randomised control trial design to evaluate the efficacy of a mobile application-based performance feedback (MABPF) intervention. The experimental approach allows for both between-group and within-group comparisons of the intervention group and the control group, providing robust data on intervention effectiveness.

The study population comprises male and female staff of Nnamdi Azikiwe University, Anambra State, Nigeria, aged 40 to 70 years who voluntarily agreed to participate. A key exclusion criterion was non-ownership of a smartphone capable of supporting the installation of the mobile application. This criterion was essential as the intervention’s core component involved the use of a smartphone application to track physical activity.

Sampling and group assignment

To ensure representativeness across the university, a proportionate stratified sampling technique was implemented. One-third of departments from 17 academic faculties spread across four locations and one non-academic unit were randomly selected using the fish-bowl method. Following recruitment, participants were randomly assigned (1:1:1) using computer-generated numbers to:

Group 1: MABPF with hourly reminders + education.

Group 2: MABPF without reminders + education.

Control: Education only.

Allocation was concealed using sequentially opened, sealed opaque envelopes after baseline data collection. While participants and staff were aware of group assignments due to the intervention’s nature, outcome assessors were blinded during data analysis. The participants were tasked to maintain their protocol, as there would be room for exposure to the main intervention post-study.

Sample size determination

Sample size was calculated using G*Power (v3.1.9.7) to detect a 0.15 effect size with 95% power at a 5% significance level. To account for attrition, the target sample was increased by 20% to 170. The final sample was 150 (88% retention), with dropouts due to smartphone incompatibility (n = 9) and withdrawal (n = 11) as depicted in Fig. 1.

Fig. 1.

Study sampling and randomisation flowchart

Outcome measures

The primary outcome was objectively measured step counts. Secondary outcomes were blood pressure, adiposity indices (body mass index [BMI], waist circumference [WC], waist-to-hip ratio [WHR], percent body fat [%BF], and visceral fat [VF]), lipid profile, blood glucose, and stress levels.

Data collection procedures

Ethical approval was obtained from the Nnamdi Azikiwe University Teaching Hospital Health Research Ethics Committee (NAUTH/CS/66/VOL.16/VER.3/33/2023/83), Nnewi. The trial was prospectively registered with the Pan African Clinical Trial Registry (PACTR202405808298188, 24 May 2024) and conducted in accordance with CONSORT guidelines. Informed consent was obtained electronically via a Google Form from all participants before any data collection commenced.

Baseline physical activity metrics, including weekly pedometer step counts (Omron HJ-325) and Google Fit-recorded step counts and heart points, were collected from all participants following group allocation but prior to any intervention, serving as the pre-intervention reference point for subsequent comparisons. Thereafter, data collection was conducted at baseline and at 12 weeks. Variables assessed comprised sociodemographic characteristics, physiological parameters, lifestyle factors, and physical activity metrics. Sociodemographic data captured participants’ age, sex, marital status, educational level, years of employment, and staff category. Blood pressure, height, weight, BMI, stress, smoking habits, alcohol use, and PA were assessed using the validated instruments and established procedures described by Chukwuemeka et al. [20]; waist circumference and WHR were assessed following Maruf et al. [21]; and blood lipid profile and blood glucose procedures followed those outlined by Chukwuemeka [3].

Percent body fat and visceral fat were measured using an Omron body composition monitor (Model BF-511) employing bioelectrical impedance analysis. Participants stood barefoot on the unit with weight equally distributed, eyes forward, and arms elevated horizontally with elbows extended at 90° to the body whilst firmly gripping the electrodes. Age, height, and sex data were entered prior to measurement, and %BF and VF values were recorded to the nearest whole number once the participant had remained still until the assessment was complete.

The MABPF tool used was the Google Fit application, a free activity tracker operating in the background of Android devices that automatically records walking, running, and cycling, and from which step counts and heart points can be extracted on a daily or weekly basis [9]. This application has been validated as a reliable instrument for step count assessment [22–24]. Objective PA was measured concurrently using both the Omron HJ-325 pedometer, which served as the validated reference instrument for step count quantification across all groups, and the Google Fit application, which functioned as the performance feedback tool for intervention participants. The Omron pedometer passively records steps as a body-worn device, whereas Google Fit additionally enables real-time self-monitoring and goal feedback.

Experimental protocol

For Intervention Groups 1 (n = 52) and 2 (n = 50), the Google Fit application was installed on participants’ smartphones, and they received training on daily PA monitoring. A daily target of 10,000 steps and 150 heart points was established. The 10,000-step target is supported by evidence associating this level of ambulatory activity with reductions in cardiovascular and all-cause mortality risk [25], and the 150 heart-point target corresponds directly to the WHO-recommended minimum of 150 min of moderate-intensity PA per week, operationalised within Google Fit where each minute of moderate activity accrues one heart point [9]. Participants were provided with a phone pouch with straps to keep their devices on their person throughout the day, reducing unaccounted steps when phones might otherwise be left on tables or in bags. Step counts were recorded every Monday by the research team directly from participants’ phones for 12 consecutive weeks (Monday to Sunday). Participants were also reminded via SMS at least three times weekly to check their feedback each evening to assess whether they had met the daily target.

Both intervention groups received educational materials, including a pamphlet and audio-visual recordings covering PA promotion and cardiovascular risk prevention. Intervention Group 1 received an additional component: the hourly reminder application (v3.5.5, Gitlab) installed on their phones, which prompted them every hour from 08:00 to 21:00 with the message: “Please, stand up and walk until you have an additional 850 steps recorded in your Google Fit app.”

The control group (n = 48) received only the educational pamphlet and audio-visual materials on PA promotion and cardiovascular risk prevention. Their PA levels and step counts were assessed weekly over the same 12-week period.

Data analysis

Data analysis was conducted using SPSS version 25.0 (IBM Corp., Chicago, IL, USA) with intention-to-treat analysis applied to account for loss to follow-up. Descriptive statistics summarised the data; categorical variables were presented as frequencies and proportions, and continuous variables as means with standard deviations (SD). The Shapiro-Wilk test confirmed non-normal data distribution (p < 0.05); non-parametric tests were therefore employed throughout. Within-group comparisons across the three time points (baseline, 6 weeks, 12 weeks) were assessed using Friedman ANOVA, with Wilcoxon Signed-Rank tests for pairwise within-group differences. Between-group comparisons were assessed using Quade’s ANCOVA, with Mann-Whitney U tests for pairwise between-group differences. The significance level was set at α = 0.05 for all analyses.

Results

Socio-demographic characteristics of the participants

The demographic and health profile of 150 staff members across three campuses: Agulu (15.5%), Awka (52%), and Nnewi (32.7%). A higher proportion of non-academic staff (69.4%) was noted. Bachelor’s degrees (38.7%), doctoral degrees (22%), and diplomas (16.7%) were the major qualifications, while O’level was the lowest (1.3%). They were predominantly females (64.7%) with a mean age of 47.11 ± 6.46 years.

The lipid profile in Table 1 shows that 47.3% had high total cholesterol, 3.3% had elevated triglycerides, and 93.3% had low HDL cholesterol levels. 28% and 24.6% of the participants have systolic and diastolic hypertension, respectively. Also, anthropometric measurements presented that the majority of the participants had high WC (76%), WHR (58.7%), body fat (71.3%) and visceral fat (62.7%). In addition, the majority of participants were overweight (32.7%) and obese (20%).

Table 1.

Participants’ baseline CVRF categories

Variables	n	Categories	Frequency	percentage
Observed Lipid and blood sugar status
Total cholesterol	150	Normal	79	52.7
		High	71	47.3
Triglyceride	150	Normal	145	96.7
		High	5	3.3
High-density lipoprotein	150	Low	140	93.3
		Normal	10	6.7
Low-density lipoprotein	150	Normal	110	73.3
		High	40	26.7
Blood sugar	150	Low	0	0
		Normal	102	68
		High	48	32
Observed Blood-pressure
Systolic blood pressure	150	Normal	58	38.7
		Pre-Hypertension	50	33.3
		Stage 1 Hypertension	34	22.7
		Stage 2 Hypertension	8	5.3
Diastolic blood pressure	150	Normal	61	40.7
		Pre-Hypertension	52	34.7
		Stage 1 Hypertension	23	15.3
		Stage 2 Hypertension	14	9.3
Blood Pressure	150	Normal	94	62.7
		Hypertension	56	37.3
Adiposity statuses
Waist circumference	150	Normal	36	24
		High	114	76
Waist-to-hip ratio	150	Normal	62	41.3
		High	88	58.7
Body Mass Index	150	Underweight	4	2.6
		Normal Weight	67	44.7
		Overweight	49	32.7
		Obesity	30	20
Percent body fat	150	Low	7	4.7
		Normal	36	24
		High	107	71.3
Visceral fat	150	Normal	56	37.3
		High	94	62.7
Smoking status	150	No	138	92
		Yes	12	8
Alcohol consumption	150	No	99	66
		Yes	51	34
Stress	150	Low	8	5
		Moderate	87	58
		High	55	37

Cutoffs: Total cholesterol: normal < 200 mg/dL, borderline high 200–239 mg/dL, high >/=240 mg/dL; HDL: low < 40 mg/dL (men), < 50 mg/dL (women); Triglycerides: normal < 150 mg/dL, elevated 150–199 mg/dL, high >/=200 mg/dL; LDL: optimal < 100 mg/dL, borderline high 130–159 mg/dL, high >/=160 mg/dL; Blood glucose: normal fasting < 100 mg/dL, pre-diabetes 100–125 mg/dL, diabetes >/=126 mg/dL; Systolic BP: normal < 120 mmHg, elevated 120–129 mmHg, hypertension stage 1: 130–139 mmHg, stage 2: >/=140 mmHg; Diastolic BP: normal < 80 mmHg, hypertension stage 1: 80–89 mmHg, stage 2: >/=90 mmHg

Additionally, the smoking prevalence is relatively low (8%), and alcohol consumption is more common (34%). Stress was common, with 58% experiencing moderate stress and 37.3% reporting high stress. The objective measure of PA showed a median pedometer step of 17,731/week (1000–32992 range), Google Fit heart point 0/week (0–11 range), and steps 12,343/week (0-25164 range), respectively. The distribution of alcohol and smoking across study groups at baseline showed no statistically significant differences (p = 0.25–0.99) as shown in Table 2.

Table 2.

Baseline PA (weekly) behaviour, alcohol use and smoking status of the participants

Alcohol use and smoking status distribution (count (%)
	Intervention group 1	Intervention group 2	Control group	X2			p-value
Alcohol use
No	33(64.7)	31(64.6)	31(66)	66.0	64.6	64.7	0.988
Yes	18(35.3)	17(35.4)	16(34)	34.0	35.4	35.3
Smoking status
No	49(94.2)	44(88)	45(93.8)	93.8	88.0	94.2	0.441
Yes	3(5.8)	6(12)	3(6.3)	6.3	12.0	5.8

Table 3 presents baseline descriptive statistics of continuous variables. The mean age of participants was 47.11 ± 6.46 years, with years of working experience averaging 10.51 ± 7.77 years. Notable metabolic parameters included mean total cholesterol of 189.91 ± 33.94 mg/dl, mean blood sugar of 122.33 ± 36.66 mg/dl, and mean blood pressure readings of 128.19/83.01 ± 19.48/12.35 mmHg. The anthropometric measurements revealed a mean BMI of 26.22 ± 6.53 and a mean percent body fat of 38.92 ± 10.29. Particularly significant was the visceral fat measurement, which showed statistically significant differences across study groups (p = 0.001).

Table 3.

Baseline mean distribution of continuous variables

Variables (n = 150)	Mean	Standard deviation	Study Groups Z-scores			p-value
			Intervention group 1	Intervention group 2	Control group
Age	47.11	6.46	68.15	72.00	86.47	0.09
YOE	10.51	7.77	75.92	80.55	69.42	0.45
Stress	12.13	4.37	71.31	73.28	81.29	0.49
TC	189.91	33.94	73.68	77.77	73.41	0.85
T	131.26	32.36	72.46	80.25	72.04	0.56
HDL	32.68	7.77	75.23	71.10	79.16	0.66
LDL	129.38	33.16	76.1	77.75	70.58	0.70
Blood sugar	122.33	36.66	66.88	73.00	86.83	0.07
SBP	128.19	19.48	71.64	75.20	78.73	0.72
DBP	83.01	12.35	75.65	74.91	74.33	0.99
WC	95.06	12.71	83.21	72.48	68.19	0.20
WHR	0.87	0.09	79.92	76.87	67.09	0.32
Height	162.27	13.03	83.07	72.22	68.66	0.22
Weight	84.53	17.58	80.97	76.57	66.19	0.23
BMI	26.22	6.53	78.66	75.18	70.49	0.65
%BF	38.92	10.29	73.82	74.81	76.60	0.95
VF	11.71	5.058	92.95	68.71	60.99	0.01*
Pedometer steps	17,731m	1000-81415R	61.11	69.70	63.08	0.25

YOE: years of employment; CVRF: cardiovascular risk factors; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; BMI: Body mass index; WC: Waist circumference; WHR: Waist−to−hip ratio; PBF: Percent body fat; VF: Visceral fat; TC: Total cholesterol; T: Triglyceride; HDL: High−density lipoprotein; LDL: Low−density lipoprotein; BG: Blood glucose; m: Median; R: Range

^*Significant at p<0.05

Within-group comparisons

The effects of the MABPF with hourly reminder intervention protocol on the CVRF showed significant improvements were observed in DBP, BMI, WC, WHR, PBF, and VF (all p < 0.05). Most improvements were notable when comparing the baseline to 12 weeks (Table 5). Systolic blood pressure showed no significant changes (p = 0.796), as shown in Table 4. A substantial decrease in stress levels across all time points. Also, improvements in total cholesterol (w = 3.189; p = 0.047) and HDL cholesterol levels (w = 6.818; p = 0.009) were noted. Moreover, there was a significant reduction in blood glucose level (w = 9.981, p < 0.001).

Table 5.

Post-hoc analysis showing the periods that accounted for the difference in CVRF among the different intervention group one participants

Movement	Time	W(p-value)
		Intervention group 1	Intervention group 2	Control group
DBP	Baseline vs. 6th week	-0.431(0.620)
	Baseline vs. 12th week	-2.226(0.033*)
	6th week vs. 12th week	-0.514(0.653)
BMI	Baseline vs. 6th week	-0.277(0.782)
	Baseline vs. 12th week	-4.921(< 0.001*)
	6th week vs. 12th week	-1.819(0.069)
WC	Baseline vs. 6th week	-2.371(0.018*)	-1.448(0.147)
	Baseline vs. 12th week	-5.844(< 0.001*)	-5.006(< 0.001*)
	6th week vs. 12th week	-0.183(0.855)	-5.790(< 0.001*)
WHR	Baseline vs. 6th week	-0.747(0.455)
	Baseline vs. 12th week	-2.529(0.011*)
	6th week vs. 12th week	-0.058(0.854)
PBF	Baseline vs. 6th week	-0.909(0.373)
	Baseline vs. 12th week	-3.492(< 0.001*)
	6th week vs. 12th week	-0.671(0.502)
VF	Baseline vs. 6th week	-3.100(0.002*)	-2.831(0.005*)
	Baseline vs. 12th week	-5.536(< 0.001*)	-4.156(< 0.001*)
	6th week vs. 12th week	-0.198(0.843)	-2.670(0.807)
Stress	Baseline vs. 6th week	-4.312 (< 0.001*)	-2.592(0.010*)	-0.847(0.397)
	Baseline vs. 12th week	-6.387(< 0.001*)	-6.281(< 0.001*)	-4.019(< 0.001*)
	6th week vs. 12th week	-4.927(< 0.001*)	-5.120(< 0.001*)	-2.004(0.045*)
Pedometer Step counts	Baseline vs. 6th week	-4.517(< 0.001*)	-1.814(0.070)	-1.779(0.075)
	Baseline vs. 12th week	-5.764(< 0.001*)	-3.745(0.001*)	-2.847(0.004*)
	6th week vs. 12th week	-4.175(< 0.001*)	-2.301(0.023*)	-1.932(0.053)

DBP: Diastolic blood pressure; BMI: Body mass index; WC: Waist circumference; WHR: Waist−to−hip ratio; PBF: Percent body fat; VF: Visceral fat

^*Significant at p < 0.05

Table 4.

Friedman analysis of variance and Wilcoxon sign rank test of difference assessing the effects of the different intervention protocols on the CVRF

Variables		MABPF with hourly reminder Intervention group			MABPF Intervention group					Control group
		F(p-value)		Effect size	F(p-value)			Effect size		F(p-value)	Effect size
SBP		0.457(0.796)		0.004	0.520(0.771)			0.004		0.231(0.891)	0.003
DBP		5.822(0.013*)		0.055	0.128(0.938)			0.055		1.349(0.509)	0.015
BMI		22.236(< 0.001*)		0.218	3.040(0.219)			0.218		0.188(0.910)	0.002
WC		29.772(< 0.001*)		0.295	15.877(< 0.001*)			0.295		1.593(0.451)	0.017
WHR		9.500(0.009*)		0.093	0.853(0.653)			0.093		1.083(0.582)	0.012
PBF		17.416(< 0.001*)		0.171	1.502(0.472)			0.171		5.408(0.067)	0.059
VF		23.598(< 0.001*)		0.231	19.583(< 0.001*)			0.231		1.924(0.382)	0.021
Stress		66.521(< 0.001*)		0.628	67.156(< 0.001*)			0.628		12.289(0.002*)	0.134
Smoking status		2.000(0.368)		0.021	0.600(0.741)			0.021		1.500(0.472)	0.016
Alcohol consumption		6.000(0.050)		0.064	0.839(0.657)			0.064		0.333(0.846)	0.004

	W(p-value)		Effect size			W(p-value)	Effect size		W(p-value)		Effect size
TC	3.189(0.047*)		0.060			2.373(0.123)	0.047		0.348(0.555)		0.008
T	0.167(0.683)		0.003			6.545(0.011*)	0.128		0.471(0.493)		0.010
HDL	6.818(0.009*)		0.129			1.800(0.180)	0.035		6.125(0.013*)		0.133
LDL	1.167(0.683)		0.003			1.190(0.275)	0.023		0.758(0.384)		0.016
BG	9.981(< 0.001*)		0.188			5.667(0.017*)	0.111		8.696(0.003*)		0.189
Pedometer Step counts	45.338(< 0.001*)		0.597			7.740(0.021*)	0.117		11.133(0.004*)		0.118

SBP: Systolic blood pressure; DBP: Diastolic blood pressure; BMI: Body mass index; WC: Waist circumference; WHR: Waist−to−hip ratio; PBF: Percent body fat; VF: Visceral fat; TC: Total cholesterol; T: Triglyceride; HDL: High−density lipoprotein; LDL: Low−density lipoprotein; BG: Blood glucose

^*Significant at p < 0.05

The effectiveness of MABPF intervention on CVRF as seen in Tables 4 and 5 shows that the blood pressure measures remained stable (p > 0.05). Significant improvements in waist circumference and visceral fat (p < 0.001) were noted, whereas other body composition measures (BMI, WHR, PBF) showed no significant changes. There was a significant decrease in stress levels across all time points (F = 67.156; p < 0.001). The significant improvements in triglycerides and blood glucose levels were observed.

The effects of the Education on CVRF and benefits of PA intervention on CVRF as depicted in Table 4 showed no significant changes in any body composition measures (BMI, WC, WHR, PBF, VF) and blood pressure (systolic and diastolic) remained stable throughout the study (p > 0.05). However, a significant reduction in stress levels were identified (p = 0.002). The difference at different time frames was between baseline and post 12th weeks (w=-4.019, p < 0.001) and 6th week and post 12th weeks (w=-2.004, p = 0.045). Also, a Significant increase in HDL cholesterol level (p = 0.013) and a decrease in blood glucose levels (0.003) were observed.

Between-group comparisons

Table 6 presents Quade’s analysis of covariance comparing three groups. The results showed that compared to the control group, the group exposed to a combination of MABPF interventions (with or without an hourly reminder) showed statistically significant improvements in the following parameters: systolic Blood Pressure (F = 5.818, p = 0.004), total Cholesterol (F = 4.976, p = 0.008), blood glucose (F = 3.354, p = 0.038), HDL (F = 3.308, p = 0.039), LDL (F = 2.850, p = 0.025), and step counts (F = 7.336, p = 0.001) Table 7 shows the MannWhitney U comparisons between the intervention and control groups. Intervention group 1 showed a significant decrease in systolic blood pressure compared to Intervention group 2 (U = 994.500 p = 0.041) and control group (U = 814.500, p = 0.003). Intervention group 1 showed significant differences in total cholesterol compared to Intervention group 2 (U = 916, p = 0.010) and control group (U = 880.500, p = 0.011). There was a significant difference in HDL between Intervention group 1 and the Intervention group 2 (U = 966, p = 0.025) and also between Intervention group 2 and the control group (U = 855.500, p = 0.014). There was a significant difference in LDL between the Intervention group 2 and the control group (U = 1044.500, p = 0.025). Blood sugar showed a significant difference between the two Intervention groups (U = 861, p = 0.019) and also between the Intervention group 2 and the control group (U = 763.500, p = 0.002).

Table 6.

Quade’s analysis of covariance assessing the effects of the intervention protocol on the cardiovascular risk factors of the participants

Variables	Mean±Standard deviation			F	P	Effect size
	Intervention group 1 (MABPF with hourly reminder)	Intervention group 2 (MABPF only)	Control group (Education on CVRF and PA benefits )
SBP	117.18 ± 23.51	126.58 ± 19.29	130.83 ± 17.53	5.818	0004*	0.073
DBP	82.02 ± 11.61	83.38 ± 13.03	84.65 ± 10.41	1.413	0.247	0.019
TC	173.04 ± 33.87	188.66 ± 26.02	183.11 ± 29.30	4.976	0.008*	0.064
T	130.44 ± 37.32	124.08 ± 22.01	127.33 ± 40.62	0.939	0.394	0.013
HDL	37.500 ± 8.90	33.900 ± 8.52	37.98 ± 8.22	3.308	0.039*	0.043
LDL	126.54 ± 30.28	129.32 ± 37.68	125.03 ± 31.75	0.362	0.025*	0.005
BS	112.60 ± 25.65	99.80 ± 32.04	113.85 ± 25.72	3.354	0.038*	0.044
BMI	24.45 ± 5.02	25.13 ± 5.56	26.13 ± 5.72	1.757	0.176	0.023
WC	90.48 ± 15.56	88.56 ± 12.29	91.04 ± 8.70	1.892	0.154	0.025
WHR	0.87 ± 0.09	0.87 ± 0.07	0.85 ± 0.07	0.831	0.438	0.011
PBF	35.64 ± 11.05	38.35 ± 12.24	40.86 ± 10.68	2.881	0.059	0.038
VF	10.56 ± 4.98	9.08 ± 4.83	9.13 ± 3.34	0.927	0.398	0.013
Stress	9.31 ± 3.91	8.74 ± 5.01	9.75 ± 3.81	1.050	0.352	0.014
Pedometer step counts	50023.04 ± 14488.77	40323.40 ± 19556.54	35451.44 ± 20454.56	7.336	0.001*	0.108

SBP: Systolic blood pressure; DBP: Diastolic blood pressure; TC: Total cholesterol; T: Triglyceride; HDL: High−density lipoprotein; LDL: Low−density lipoprotein; BS: Blood sugar; BMI: Body mass index; WC: Waist circumference; WHR: Waist−to−hip ratio; PBF: Percent body fat; VF: Visceral fat

^*Significant at p < 0.05

Table 7.

Post-hoc analysis showing the periods that accounted for the difference in the CVRF of the participants

Movement	Time	U	P-value
SBP	Intervention 1 vs. Intervention 2	994.500	0.041*
	Intervention 1 vs. Control	814.500	0.003*
	Intervention 2 vs. Control	1006.500	0.169
TC	Intervention 1 vs. Intervention 2	916.000	0.010*
	Intervention 1 vs. Control	880.500	0.011*
	Intervention 2 vs. Control	1183.000	0.904
HDL	Intervention 1 vs. Intervention 2	966.000	0.025*
	Intervention 1 vs. Control	1221.000	0.852
	Intervention 2 vs. Control	855.500	0.014*
LDL	Intervention 1 vs. Intervention 2	1185.500	0.269
	Intervention 1 vs. Control	1244.500	0.981
	Intervention 2 vs. Control	1044.500	0.025*
BS	Intervention 1 vs. Intervention 2	861.000	0.019*
	Intervention 1 vs. Control	1212.500	0.806
	Intervention 2 vs. Control	763.500	0.002*
Pedometer step counts	Intervention 1 vs. Intervention 2	721.500	0.002*
	Intervention 1 vs. Control	482.000	< 0.001*
	Intervention 2 vs. Control	889.000	0.265

SBP: Systolic blood pressure; TC: Total cholesterol; HDL: High−density lipoprotein; LDL: Low−density lipoprotein; BS: Blood sugar

^*Significant at p < 0.05

Pedometer step counts showed significant difference between the two Intervention group (U = 721.500, p = 0.002) and between Intervention group 1 and control group (U = 482, p < 0.001).

Discussion

This study was conducted amongst 150 staff members of Nnamdi Azikiwe University, recruited from three campus locations. Non-academic staff constituted 58% of participants. The primary objective was to determine whether mobile-application-based performance feedback (MABPF) influenced cardiovascular risk factors within a One Health framework. The effectiveness of MABPF intervention with hourly reminders on various cardiovascular risk factors was examined over 12 weeks. A pre- and post-intervention design was employed with measurements taken at baseline, 6 weeks, and 12 weeks. Significant positive impacts of the MABPF intervention were observed in diastolic blood pressure, body mass index, waist circumference, waist-to-hip ratio, percent body fat, and visceral fat. Post-hoc analysis demonstrated that most improvements were notable when baseline measurements were compared to the 12-week mark, suggesting a cumulative intervention effect over time. However, systolic blood pressure did not exhibit significant changes.

The clinical significance of these findings was substantial from a One Health perspective. Cardiovascular disease prevention in university staff represents a critical component of community health promotion, as workplace wellness directly impacts broader societal health outcomes [26]. The meaningful reduction in diastolic blood pressure observed could be translated to a considerable decrease in cardiovascular risk across the university community. Furthermore, simultaneous improvements in BMI, waist circumference, waist-to-hip ratio, percent body fat, and visceral fat suggested a comprehensive effect on body composition, potentially driven by increased energy expenditure and improved metabolic function due to enhanced physical activity behaviour. These findings aligned with evidence that mobile-based interventions targeting sedentary behaviour in office workers show greatest efficacy for visceral fat reduction even before changes in overall weight become apparent [27].

The absence of significant changes in systolic blood pressure, despite significant improvements in diastolic blood pressure and other cardiovascular risk factors, warrants further explanation. Evidence from a large network meta-analysis of 270 randomised controlled trials suggests that although most exercise modalities reduce both systolic and diastolic blood pressure, the magnitude of systolic reduction varies considerably by intervention type, intensity, and duration, and short-term walking-based programmes may yield more modest systolic effects [28]. The 12-week duration of this study may therefore have been insufficient to elicit detectable systolic changes, particularly given the low-to-moderate intensity of the physical activity promoted. Additionally, pre-existing comorbidities such as hypertension and diabetes, prevalent in this cohort, and possible concurrent antihypertensive medication use, which was not controlled for, may have maintained systolic values near a pharmacological ceiling, limiting detectable intervention-related change [29]. The relatively short intervention duration, baseline systolic blood pressure levels, or differential responsiveness of systolic and diastolic blood pressure to lifestyle modifications may have contributed to these findings [30, 31]. Diastolic blood pressure is generally more responsive to short-term lifestyle modifications, likely reflecting earlier improvements in peripheral vascular resistance and arterial compliance [28, 29]. Future studies could stratify analyses by comorbidity status and medication use to better disentangle these effects.

Significant improvements in total cholesterol and HDL cholesterol levels were demonstrated by the MABPF intervention, alongside significant reductions in blood glucose levels. These findings aligned with robust evidence supporting the beneficial effects of physical activity and lifestyle modifications on lipid profiles, blood glucose control, and overall cardiovascular risk [32]. The significant decrease in blood glucose levels suggested that the intervention may have been particularly beneficial for individuals with pre-diabetes or type 2 diabetes, conditions that are increasingly prevalent in workplace communities globally [33]. Several mechanisms were likely responsible for these observed improvements in metabolic markers. Physical activity is known to increase insulin sensitivity, improve lipid metabolism, and reduce inflammation, all of which contribute to improved cardiovascular health [34]. The MABPF intervention, by promoting physical activity and potentially reducing sedentary time through its feedback mechanisms, may have triggered these beneficial physiological adaptations. From a One Health standpoint, these individual health improvements contribute to reduced healthcare burden and enhanced community resilience [16].

A substantial impact in reducing stress levels across all time points was also demonstrated by the MABPF intervention, highlighting its multifaceted effect within the One Health framework. Stress reduction in university staff has implications beyond individual health, affecting teaching quality, student wellbeing, and overall institutional effectiveness [18]. By reducing stress, the intervention may have created a positive feedback loop reinforcing healthy behaviours. Chomiuk et al. indicated that stress reduction may be a primary mediator between increased physical activity and metabolic improvements in professionals, suggesting that interventions targeting both physical activity and stress simultaneously may achieve synergistic benefits [35]. The positive effects of the MABPF intervention with hourly reminders could therefore be attributed to the combined action of behavioural and physiological mechanisms: hourly reminders served as cues to action, and the application provided real-time progress information that enhanced self-awareness and motivation, whilst the resulting increase in physical activity enhanced insulin sensitivity, reduced inflammation, and promoted favourable changes in body composition.

The effectiveness of MABPF without hourly reminders on cardiovascular risk factors also yielded mixed results, with some risk factors showing significant improvements and others remaining relatively stable. Both systolic and diastolic blood pressure remained stable throughout the intervention period, contrasting with previous studies showing that lifestyle interventions delivered via mobile applications could reduce blood pressure [36, 37]. Significant improvements in waist circumference and visceral fat were nonetheless observed, aligning with research indicating that mHealth interventions can effectively reduce abdominal adiposity [38]. The reduction in these measures was particularly noteworthy, as waist circumference and visceral fat are strong independent predictors of cardiovascular disease and metabolic syndrome [39]. The absence of concomitant changes in BMI or waist-to-hip ratio suggested that the intervention may have specifically targeted abdominal fat stores through increased incidental physical activity. Significant improvements in triglyceride and blood glucose levels were also demonstrated [40], and stress levels declined significantly across all time points, consistent with research supporting the role of mHealth interventions in reducing stress and improving mental wellbeing [41].

The education-only control intervention presented a mixed picture. No significant changes were observed in body composition measures or blood pressure, contrasting with evidence for the effectiveness of lifestyle interventions in improving these outcomes [42]. The lack of significant changes likely reflected the nature of an exclusively educational approach, which may increase awareness without producing the consistent behavioural activation required to drive physiological change. Significant increases in HDL cholesterol and reductions in blood glucose were nonetheless observed, consistent with previous research on the beneficial effects of lifestyle education for lipid profiles and glucose metabolism [32, 43].

When compared to the control group, both MABPF intervention groups showed improvements in several key parameters. Systolic blood pressure was significantly reduced in the MABPF with hourly reminders group compared to both the MABPF without reminders group and the control group, aligning with systematic reviews of mHealth technologies for cardiovascular disease management, which report that mobile interventions are particularly effective for blood pressure management when structured prompting is incorporated [44]. The findings of this trial thus carry practical implications across all three One Health dimensions. At the human health level, the MABPF intervention produced statistically significant and clinically relevant improvements in systolic blood pressure, total cholesterol, blood glucose, and weekly step counts, collectively reducing ten-year cardiovascular disease risk. Institutionally, the delivery model integrated seamlessly with existing university infrastructure at no extra cost, as participants used their own smartphones, standard text messaging, and a free application without requiring new equipment, space, or clinical staff. Environmentally, this approach suits workplaces where conventional health promotion resources are scarce. By aligning with existing communication habits rather than requiring new systems, these results extend the evidence of Odeny et al. [45] regarding mobile-based health interventions in sub-Saharan Africa to cardiovascular risk reduction among Nigerian university staff. Consequently, MABPF offers a practical, low-cost occupational health policy for resource-limited settings by simultaneously addressing individual risk, institutional delivery, and environmental feasibility.

The differential effect of the two interventions was an intriguing finding. Although the MABPF without reminders group demonstrated improvements in various parameters compared to the control, the MABPF with hourly reminders group exhibited more pronounced enhancements across multiple measures. This suggested that the addition of regular prompts significantly enhanced the effectiveness of the mobile application feedback system, and the contrasting results between the two intervention groups underscored the importance of both timing and frequency in the design of mHealth interventions.

Implications for clinical practice, public health, and one health

The trial achieved measurable improvements in diastolic blood pressure, body composition, and key metabolic markers among a working population with markedly unfavourable baseline profiles: 47.3% had elevated total cholesterol, 76% had high waist circumference, and 37.3% reported high stress. Critically, these gains required no dedicated clinical facilities or specialist personnel. The free smartphone application, standard text messaging, and existing institutional communications infrastructure were sufficient, making this model highly practical for Nigerian universities where occupational health provision is minimal or absent. By proving that a low-cost digital tool can mitigate cardiovascular risks in a controlled workplace trial, this study provides direct empirical support for the World Health Organisation’s call for scalable, technology-enabled NCD prevention in sub-Saharan Africa [46].

From an institutional and community health perspective within the One Health framework, a university is an influential social environment rather than merely a workplace. Staff members occupy visible roles, modelling health behaviours that naturally propagate to students, peers, and families. Because health behaviour changes can cascade through social networks across multiple degrees of separation, the cardiovascular improvements observed here carry implications beyond the individual participants. Reducing the cardiovascular disease burden among staff also lowers absenteeism, sustains productivity, and reduces pressure on institutional health budgets, reinforcing the case for embedding preventive digital health interventions within university policy rather than leaving wellness entirely to individual initiative [47].

The scalability of the MABPF intervention is a particularly important public health consideration. Given the low cost of smartphone-based tools and the high penetration of mobile phones across urban Nigeria, this intervention has potential for large-scale deployment across university and other workplace settings without substantial infrastructure investment [48]. At the population level, even modest reductions in CVRF translate to meaningful decreases in cardiovascular events, hospitalisations, and premature mortality [49]. If sustained, the demonstrated improvements in lipid profiles, blood glucose, and diastolic blood pressure could contribute to a measurable reduction in the ten-year CVD risk of participants. These findings reinforce calls for the integration of digital health tools into national NCD prevention strategies in sub-Saharan Africa [50], and are consistent with current recommendations for comprehensive cardiovascular risk management that advocate for innovative technologies in preventive care [51, 52].

Mobile health interventions have been shown to be effective in supporting behaviour change, which is essential for long-term adherence to lifestyle modifications that promote cardiovascular health [53]. The use of frequent mobile prompts helped sustain user engagement and encouraged consistent physical activity. The workplace setting represents a critical intervention point within the One Health framework, given that university staff serve as role models within their communities and their health behaviours influence students, families, and broader social networks [19]. The observed improvements in cardiovascular risk factors among university staff could therefore have cascading effects on community health outcomes, and healthier university staff also contribute to improved institutional productivity and reduced healthcare costs [47].

Limitations and future directions

Several limitations warrant consideration within the One Health context. The absence of long-term follow-up data limits assessment of the sustainability of observed improvements. Evidence suggests that the benefits of mHealth interventions may diminish over time if user engagement strategies are not maintained [54], and future studies should therefore incorporate extended follow-up periods. The single-centre design limits generalisability across diverse populations and settings; future research with larger, more diverse samples drawn from multiple university populations and other workplace contexts would strengthen external validity [17, 55]. Environmental factors potentially moderating intervention effectiveness, including air quality, built environment characteristics, access to green spaces, and neighbourhood walkability, were not comprehensively evaluated [56], and future studies should incorporate environmental assessments to better understand these contextual influences. The study also did not examine whether individual behaviour changes produced spillover effects on family members, colleagues, or community networks, which is a consideration of particular relevance within a One Health analytical framework [57]. The impact on healthcare utilisation and costs was not assessed; future studies should incorporate health economic evaluations to establish the cost-effectiveness of MABPF across different healthcare systems. Finally, cultural and contextual factors that may have influenced intervention acceptability and engagement were not thoroughly explored, and future research should examine how such factors shape outcomes across diverse populations [58].

Conclusion

This trial demonstrates that MABPF with hourly reminders produces significant, clinically meaningful improvements in CVRF among Nigerian university staff over 12 weeks. Analysed through the One Health framework, the intervention simultaneously addressed individual metabolic risk, utilised existing university infrastructure at zero additional cost, and proved feasible within a resource-limited environment lacking conventional health facilities. The superior effectiveness of the reminder condition over the application-only group highlights how structured prompting successfully counters the effects of prolonged sedentary behaviour in desk-based roles. The low-cost, adaptable nature of this intervention makes it particularly well-suited to resource-limited settings where comprehensive prevention programmes are otherwise unfeasible, and the university workplace represents a strategically important intervention point, given that staff health behaviours can cascade to students, families, and broader social networks. The observed reductions in stress, improved metabolic markers, and favourable body composition changes collectively demonstrate that MABPF addresses multiple cardiovascular risk pathways. These findings strengthen the evidence base for digital health technologies in resource-limited settings, supporting the integration of individual behaviour change within broader institutional health systems as a scalable and cost-effective approach to reducing the rising burden of CVD across sub-Saharan Africa.

Author contributions

Conceptualisation: U.M.C., A.C.A., F.S.N., I.A.A., B.E.L., I.U.N., E.S.N., G.F., F.A.M., C.E-D., B.B.; Methodology: U.M.C., A.C.A., G.F., F.S.N., J.L, C.E-D.; Investigation: U.M.C., A.C.A., I.A.A.; Formal Analysis: U.M.C., B.E.L., F.A.M.; Resources: U.M.C., A.C.A., I.A.A., C.E-D., G.F., B.B.; Data Curation: B.E.L., U.M.C., F.A.M., I.A.A.; Writing—Original Draft: U.M.C., A.C.A.; Writing—Review and Editing: All authors; Visualisation: U.M.C., F.S.N., G.F., F.A.M.; Supervision: G.F., F.S.N., I.U.N., E.S.N., C.E.D., J.L., B.B.; Project Administration: U.M.C., A.C.A., F.A.M.; Funding Acquisition: U.M.C., F.A.M. All authors have read and approved the final version of the manuscript.

Funding

This study was funded by the Science for Africa Foundation through the Developing Excellence in Leadership, Training and Science in Africa (DELTAS Africa) programme (Afrique One-ASPIRE, Del-15-008 and Afrique One-REACH, Del-22-011) with support from the Wellcome Trust and the UK Foreign, Commonwealth & Development Office. The study is part of the EDCPT2 programme supported by the European Union. The funders had no role in study design, data collection, analysis, interpretation, or manuscript preparation.

Data availability

The datasets supporting the conclusions of this article are available from the corresponding author upon reasonable request and following appropriate ethical approval for data sharing.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the principles of the Declaration of Helsinki (revised 2013) and Good Clinical Practice guidelines. Ethical approval was obtained from the Nnamdi Azikiwe University Teaching Hospital Health Research Ethics Committee, Nnewi, Nigeria (Approval Number: NAUTH/CS/66/VOL.16/VER.3/33/2023/83) prior to participant recruitment. The trial was prospectively registered with the Pan African Clinical Trial Registry (Trial Number: PACTR202405808298188, registered 24 May 2024). Written informed consent was obtained electronically from all participants via a structured Google Form before any data collection commenced. Each participant was fully briefed on the study objectives, procedures, potential risks, anticipated benefits, and their right to withdraw at any point without consequence.

Consent for publication

Not applicable. This manuscript contains no individually identifiable data, images, or recordings relating to any participant.

Competing interests

The authors declare no competing interests.