Ambulatory blood pressure monitoring phenotypes and circadian blood pressure patterns among adults referred for diagnostic evaluation at Aga Khan University Hospital, Nairobi, Kenya: a retrospective cross-sectional study

¹Department of Internal Medicine, Aga Khan University Hospital, Nairobi, Kenya, ²Department of Population Health, Aga Khan University Hospital, Nairobi, Kenya

^&Corresponding author
Nadeem Kassam, Department of Internal Medicine, Aga Khan University Hospital, Nairobi, Kenya

Introduction    

Hypertension continues to be the primary modifiable risk factor for cardiovascular illness and death globally, significantly adding to the burden of atherosclerotic cardiovascular disease (ASCVD) [1]. This challenge is particularly pronounced in middle- and low-income regions, where detection and control rates are significantly lower than those in high-income countries [2]. In sub-Saharan Africa (SSA), hypertension prevalence has been increasing quickly, with recent estimates indicating that almost 50% of adults are hypertensive [3]. Projections suggest that the affected population will grow from 130 million in 2010 to over 215 million by 2030 [4]. Importantly, SSA bears the highest rates of hypertension-related mortality globally, reflecting gaps in both diagnosis and effective management [5]. Kenya mirrors this regional trend, with hypertension emerging as one of the most pressing non-communicable diseases. National surveys have documented an increasing prevalence across both urban and rural populations, yet awareness, diagnosis, and treatment rates remain low [6,7]. Conventional office blood pressure measurements, which are prevalent in diagnostics, are hindered by issues like white coat hypertension, masked hypertension, and inaccuracy and reproducibility [8,9]. These limitations contribute to both underdiagnosis and overdiagnosis, ultimately impeding appropriate management and prevention of cardiovascular disease.

Ambulatory blood pressure monitoring offers a solution by providing a 24-hour blood pressure profile that captures variability across both day and night [10,11]. Unlike single office measurements, ABPM identifies abnormal circadian patterns, including nocturnal hypertension and non-dipping or reverse-dipping profiles, which are strong and independent predictors of adverse cardiovascular and renal outcomes [12,13]. The aforementioned enhances diagnostic accuracy and helps clinicians tailor treatment strategies more effectively, including lifestyle interventions and medication. Despite its proven value, ABPM remains underutilized in SSA, with fewer than 10% of health facilities having access to this technology [14]. Barriers include high costs, limited availability, and a lack of awareness among clinicians [15,16]. Consequently, data on ABPM use and hypertension phenotypes in SSA, and particularly in Kenya, remain scarce. Given these gaps, we analyzed ABPM data from the Aga Khan University Hospital, Nairobi. The primary objectives of this study were to report patient demographics, cardiovascular risk profiles, and blood pressure trends detected by ABPM, with a special emphasis on dipping status. The purpose of this study was to evaluate the efficiency of ambulatory blood pressure monitoring (ABPM) in our setting and to examine blood pressure circadian rhythms. We investigated the incidence of dipping anomalies and nocturnal hypertension and their potential factors associated with increased cardiovascular risk.

Methods     

Study design and duration: this study employed a retrospective, cross-sectional design and reviewed data collected over 30 months, from January 2022 through June 2024.

Study area and population: the research was conducted at Aga Khan University Hospital in Nairobi (AKUHN), a tertiary referral and teaching hospital with a specialized outpatient cardiology unit that routinely offers Ambulatory Blood Pressure Monitoring (ABPM) as part of routine patient management.

Inclusion criteria: the study population comprised all adults (≥18 years) who underwent ABPM between January 2022 and June 2024, patients exhibiting elevated office blood pressure readings without a previous diagnosis of hypertension.

Exclusion criteria: patients with previously established hypertension already on treatment, and those with incomplete or technically invalid ABPM recordings, were excluded. The study, therefore, focused on individuals referred for diagnostic clarification, including those with suspected white coat hypertension or inconsistent office readings.

Variables

Independent variables: Age; ii) sex; iii) BMI (continuous); iv) diabetes mellitus (yes/no); v) HbA1c; vi) triglycerides (TG); vii) total cholesterol (TC); Viii) LDL cholesterol; ix) non-HDL cholesterol dependent variables; x) adverse circadian blood pressure pattern (non-dipping/reverse dipping status).

sample size: all eligible patients who met the inclusion criteria during the specified study period were included. A formal sample size calculation was not undertaken because the study comprised a complete review of all available ABPM records during this timeframe.

Blood pressure measurements: office blood pressure was measured on the left arm using a mercury sphygmomanometer after the participant had rested for five minutes in a seated position, with appropriate posture and cuff size. Elevated readings were repeated after five minutes and recorded. Ambulatory blood pressure monitoring (ABPM) was initiated following cardiology review using a validated automated device at Aga Khan University Hospital, Nairobi. Measurements were obtained every 15 minutes during daytime (07:00-22:00) and every 30 minutes overnight (22:00-07:00). Only recordings with at least 20 daytime and 7-10 nighttime measurements were included in the analysis [17].

Data collection: data were extracted from institutional ABPM logs and the electronic medical record system. Variables collected included demographic information, relevant clinical characteristics such as family history of hypertension or cardiovascular disease, and cardiovascular risk factors including diabetes mellitus, dyslipidemia, and smoking status. Ambulatory blood pressure monitoring-derived variables included mean systolic and diastolic blood pressures over 24-hour, daytime, and nighttime periods, nocturnal dipping status, and circadian blood pressure patterns. All ABPM procedures were performed using validated devices in routine clinical use at AKUHN. Patients were advised to continue their usual daily activities while avoiding strenuous exertion and to maintain a sleep diary to facilitate accurate classification of daytime and nighttime readings. Recordings were considered valid if at least 70% of expected measurements were successful, including a minimum of 20 daytime and 7 nighttime readings.

Definitions for blood pressure patterns: the institution used the AHA/ACC 2017 criteria [18] to interpret ABPM results, as the study and data extraction were conducted before the release of the updated AHA/ACC 2025 hypertension guidelines [19]. Hypertension on ABPM was considered present if any of the following criteria were met: mean 24-hour blood pressure ≥130/80 mmHg, mean daytime (awake) blood pressure ≥135/85 mmHg, or mean nighttime (asleep) blood pressure ≥120/70 mmHg. Circadian blood pressure patterns were also classified, with a normal dipping pattern defined as a nocturnal systolic blood pressure fall of 10-20% compared to daytime levels. Non-dipping was defined as a nocturnal fall of <10%, reverse dipping as a rise in nighttime blood pressure compared to daytime, and extreme dipping as a fall of >20% [20].

Data handling and analysis:data were analyzed using SPSS version 20 (IBM Corp., Armonk, NY). Continuous variables were summarized using means with standard deviations or medians with interquartile ranges, as appropriate. Categorical variables were presented as frequencies and proportions. Comparisons between categorical variables were performed using chi-square or Fisher´s exact tests, while continuous variables were compared using Student´s t-test. Multivariable logistic regression analysis was conducted to identify independent predictors of adverse circadian blood pressure patterns, specifically non-dipping and reverse-dipping phenotypes, which are associated with increased atherosclerotic cardiovascular disease risk. Statistical significance was defined as a two-sided p-value <0.05.

Ethical considerations: ethical approval for this study was obtained under the umbrella of a larger hypertension research programme being conducted by the section. The approval covered all hypertension-related projects. An addendum was subsequently submitted and approved to incorporate ambulatory ABPM data. All sectional, departmental, and institutional ethical requirements were addressed collectively as part of this broader approval framework. The study received approval from the ISERC (Ref: 2025/ISERC-004 [v2]) and a research permit from NACOSTI (Permit No. NACOSTI/P/25/4181175).

Results     

A total of 180 patients were included in the final dataset. The median age of the cohort was 47 years (interquartile range [IQR] 38-62), with the majority being over 40 years of age (n = 124, 68.9%) and of African descent (n =149, 83.7%). More than half of the participants were male (n= 95, 52.8%). The median body mass index (BMI) was 27.5 kg/m² (range: 24.6-30.4), as detailed in Table 1. Among the 180 patients, approximately half were diabetic (n = 78, 46.2%), with the majority of them receiving oral hypoglycemic agents (n = 15, 68.1%). The underlying comorbidities are presented in Table 2 below. Of the 180 participants, 122 had their lipid profiles recorded during their clinic visit. The median total cholesterol was 4.9 mmol/L (range: 4.2-5.7), and the median LDL cholesterol was 3.4 mmol/L (range: 2.7-4.1). Among the 93 participants with available HbA1c results, the median value was 5.9% (range: 5.6-6.2). Detailed laboratory parameters for the cohort are presented in Table 3 below.

The overall mean office systolic blood pressure was 143.1 ± 20.1 mmHg, with a mean diastolic pressure of 82.4 ± 12.5 mmHg. When analyzed by dipping pattern, both normal dippers and extreme dippers exhibited higher office systolic values (151.3 ± 25.9 mmHg and 157.2 ± 28.2 mmHg, respectively), whereas reverse dippers had the lowest office systolic reading at 139.6 ± 18.9 mmHg. A similar pattern was observed for diastolic pressure: Normal dippers recorded the highest mean value (87.1 ± 13.1 mmHg), while reverse dippers had the lowest (79.2 ± 13.6 mmHg). Reverse dippers recorded the highest night-time systolic and diastolic pressures (142.3 ± 15.5 mmHg and 83.7 ± 11.1 mmHg, respectively). Detailed ambulatory blood pressure monitoring (ABPM) findings are summarized in Table 4 below.

A substantial proportion of participants (95.0%, n= 171) were diagnosed with hypertension based on ambulatory blood pressure monitoring (ABPM). The lowest prevalence was observed among dippers (n = 29, 85.3%). Interestingly, night-time hypertension was more common than daytime hypertension across all subgroups, affecting 84.4% of the overall cohort. Reverse dippers showed the highest prevalence of night-time hypertension (92.5%), followed by non-dippers (84.9%), highlighting the significant burden of elevated nocturnal blood pressure in these patterns. The description is noted in Table 5 below. Given the limited sample size, for analysis, normal dippers and extreme dippers were combined into a single dipping group. In contrast, non-dippers and reverse dippers were combined into a single non-dipping group. Among the 180 patients, 139 (77%) were classified as non-dippers, whereas 41 (23%) were classified as dippers. Additional details are provided in Table 6 below.

We conducted logistic regression analysis to identify factors linked to non-dipping status, a phenotype associated with increased cardiovascular risk. In the multivariable model, male gender (aOR 1.15, 95% CI: 0.30-4.36, P=0.8), diabetes mellitus (aOR 1.64, 95% CI: 0.19-35.50, P=0.6), higher BMI (aOR 1.01, 95% CI: 0.93-1.14, P=0.8), higher HbA1c (aOR 1.07, 95% CI: 0.68-2.12, P=0.8), as well as elevated triglycerides (aOR 1.36, 95% CI: 0.65-3.92), total cholesterol (aOR 1.84, 95% CI: 0.27-16.91, P=0.5), and LDL cholesterol (aOR 1.57, 95% CI: 0.38-7.25.P=0.5) were associated with increased odds of non-dipping. However, none of these associations reached statistical significance; the study may have been underpowered to detect definitive relationships, as seen in Table 7 below.

Discussion     

Diagnostic yield of ambulatory blood pressure monitoring: our study demonstrates a very high diagnostic yield of ABPM, with 95% of patients tested meeting the criteria for hypertension diagnosis. This highlights the importance of careful patient selection and referral practices, suggesting that clinicians are appropriately identifying individuals at risk where ABPM provides incremental diagnostic value. The high proportion of positive results also reflects the significant burden of hypertension that may otherwise be under-recognized with conventional office blood pressure measurements alone.

Nocturnal hypertension and Abnormal circadian blood pressure patterns: an important finding of our study was the high burden of nocturnal hypertension, with 152 of the 175 positive cases meeting diagnostic criteria. This highlights the importance of ABPM in identifying a silent yet clinically significant phenotype strongly associated with adverse cardiovascular outcomes. We hypothesize our findings based on the fact that he majority of our cohort was overweight, suggesting that excess weight may contribute to both nocturnal hypertension and abnormal dipping patterns. Although BMI showed an increased odds ratio without reaching statistical significance, the trend is consistent with prior evidence implicating adiposity in altered circadian blood pressure regulation and sympathetic overactivity [21]. Our retrospective design limited our assessment of waist circumference; however, central obesity is well recognized as a driver of nocturnal hypertension and obstructive sleep apnea (OSA). Given the bidirectional relationship between obesity and OSA, and their shared impact on nocturnal hypertension and non-dipping [22], future prospective studies should evaluate this pathway more systematically. Clinicians should maintain a high index of suspicion for OSA in patients with nocturnal hypertension and consider incorporating validated tools such as Stop-Bang [23] into routine assessment. Additionally, more than three-quarters of patients in our cohort were non-dippers or reverse dippers, a pattern frequently coexisting with nocturnal hypertension and associated with left ventricular hypertrophy, ischemic heart disease, and stroke. Their predominance in this overweight population reinforces the interplay between adiposity, autonomic dysregulation, and sleep-disordered breathing, highlighting the importance of targeted screening and aggressive risk factor modification.

Ambulatory blood pressure monitoring evidence from sub-Saharan Africa and global comparisons: the paucity of ABPM data in sub-Saharan Africa makes our findings particularly important. In Kenya, previous community-based evidence from Kilifi has demonstrated that clinic readings frequently overestimate the prevalence of hypertension, highlighting the diagnostic value of ABPM [24]. Comparable results from Tanzania revealed a high burden of non-dipping and nocturnal hypertension [25], while Ugandan cohorts [26], though reporting lower nocturnal hypertension, highlighted a striking prevalence of white coat and masked hypertension. Beyond East Africa, studies from Nigeria [27], South Africa [28], and Egypt [29] similarly demonstrate high rates of non-dipping and nocturnal hypertension across diverse populations, reinforcing that these abnormal circadian patterns are widespread on the continent. These findings are parallel to large European [20] and North American cohorts [30], where non-dipping status and elevated night-time blood pressure are consistently shown to predict cardiovascular and renal morbidity, independent of daytime or office measurements. Taken together, the convergence of African and Western evidence supports the pathophysiological importance of nocturnal hypertension and non-dipping, likely driven by obesity, autonomic dysregulation, sleep-disordered breathing, and high salt intake. In context, our results, showing an overweight population and a heavy burden of nocturnal hypertension, align closely with this global literature and argue strongly for broader adoption of ABPM in Kenya, both in public and private settings, to enable early detection and tailored risk reduction. We did not identify significant clinical predictors of dipping versus non-dipping status in our cohort. Interestingly, non-dippers had lower cholesterol and triglyceride levels than dippers, a paradox also noted in prior reviews [29]. This may reflect sample size, unmeasured lifestyle or genetic factors, and suggests that dipping status is influenced by mechanisms beyond conventional metabolic markers, including autonomic dysfunction, renal sodium handling, vascular stiffness, sleep-disordered breathing, and chronobiological factors. Larger, more diverse studies with robust biomarkers and physiologic measures are needed to clarify these associations.

Limitations: this study has limitations: it was conducted at a single tertiary care center, limiting generalizability. The descriptive, cross-sectional design prevents causal inferences or long-term outcome assessment. Although we explored clinical and metabolic factors related to dipping patterns, the modest sample size limited detection of subtle associations, and residual confounding may remain. Fourth, information on important covariates such as sleep quality, obstructive sleep apnea, sodium intake, medication timing, and autonomic function was not available, all of which may influence nocturnal blood pressure and circadian patterns. Finally, referral bias may have contributed to the very high diagnostic yield, as patients selected for ABPM were likely those already suspected to be at elevated risk. These limitations underscore the need for larger, multicenter, prospective studies that incorporate a broader range of clinical and lifestyle variables to validate and elaborate on our findings.

Future perspective: the high prevalence of nocturnal hypertension and abnormal dipping patterns observed in our cohort highlights the need for further research to better characterize the determinants and consequences of circadian blood pressure abnormalities. Future studies should be multicenter and adequately powered to explore the influence of clinical, metabolic, and lifestyle factors, as well as the role of sleep-disordered breathing and autonomic dysfunction, in shaping blood pressure profiles. Longitudinal follow-up with cardiovascular and renal outcomes will be critical to quantify the prognostic significance of these patterns in our setting. In addition, interventional studies are needed to evaluate whether strategies such as chronotherapy, sodium reduction, weight management, or targeted treatment of sleep apnea can modify dipping status and improve outcomes. Expanding access to ABPM in routine clinical practice, particularly in resource-limited settings, should also be a priority, as this would allow earlier detection of silent hypertension and facilitate more individualized risk stratification and management.

Conclusion     

Our research highlights the exceptionally high diagnostic yield of ABPM in appropriately selected patient populations, particularly emphasizing the significant prevalence of nocturnal hypertension and abnormal dipping patterns. These findings expose the limitations of office blood pressure measurements in accurately assessing cardiovascular risk and reinforce the crucial role of ABPM in diagnosis and risk stratification. Although our analysis did not identify specific clinical or metabolic predictors of abnormal circadian blood pressure variation, the high incidence of non-dipping and reverse dipping patterns in this cohort underscores the need for careful monitoring and personalized management strategies. Future prospective studies involving larger and more diverse populations, incorporating clinical, metabolic, and sleep-related variables, are needed to better understand the factors influencing nocturnal hypertension and its implications. Until such data are available, our findings strongly support routine use of ABPM in clinical practice, especially for patients at increased risk of adverse cardiovascular events.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

All authors contributed equally to the conception and design of the study, data collection, data analysis, and manuscript preparation. All authors take full responsibility for the integrity and accuracy of the work submitted. All the authors have read and agreed too the final manuscript.

Tables     

Table 1: basic demographic characteristics of the cohort referred for ABPM monitoring (N=180)

Table 2: distribution of clinical characteristics and pre-existing comorbidities among the study participants (N=180)

Table 3: laboratory profile of the study population

Table 4: office blood pressure measurements and 24-hour ambulatory blood pressure monitoring findings among study participants recruited (N=180)

Table 5: nocturnal blood pressure dipping patterns and hypertension classification among study participants (N=180)

Table 6: differences in clinical and biochemical characteristics between participants With normal dipping and non-dipping patterns (N=180)

Table 7: factors linked to non-dipping blood pressure pattern among study participants

References