Prevalence of Prehypertension in Nigeria: a Systematic Review and Meta-analysis

Background: Prehypertension is a low blood pressure state associated with an elevated risk of cardiovascular disease and the development of hypertension. Everyone is concerned about the rising global burden of hypertension and prehypertension. The goal of this study was to determine the prevalence of prehypertension in Nigeria. Methods: Studies were chosen based on predefined criteria after doing online searches of Google Scholar, PubMed, and the African Index Medicus. The meta-analysis comprised 21 studies with a total of 25839 participants that were completed between 2011 and 2021. Results: The pooled prevalence of prehypertension in Nigeria was found to be 34% (95% CI: 30%-40%), corresponding to 41.4 million adults. Males have a higher prehypertension prevalence of 39.1% (95% CI: 30.9%-47.6%) than females, with a prevalence of 28.5% (95% CI: 21.2%-36.4%). Between the sexes, there were no statistically significant differences. While the North-East had the lowest rate of prehypertension (18%), the North-West had the greatest rate of prehypertension (43%), indicating a statistically significant differential between the regions. However, when compared to metropolitan settlements (37%), rural areas have a pooled frequency of 32 percent. Conclusion: There were no statistically significant differences in gender or between rural and urban communities. There was, nevertheless, a statistically significant difference across the regions. Prehypertension is common in Nigeria, and it indicates a future risk of hypertension and other cardiovascular disorders.


Introduction
Blood pressure that is abnormally high is known as hypertension. From a systolic blood pressure (SBP) of 115 mmHg and a diastolic blood pressure (DBP) of 75 mmHg, the risk of cardiovascular disease grows in a log-linear pattern, with the risk of cardiovascularrelated death doubling every 20 mmHg and 10 mmHg increase in SBP and DBP, respectively (1). Despite the constant risk of rising blood pressure, identifying a cut-off number is critical for clinical and public health interventions to establish a threshold of action (2). This was described as an SBP of 140 mmHg and/or a DBP of 90 mmHg. Given the long-term relationship between blood pressure and cardiovascular risk, it's only natural to establish additional low-blood pressure thresholds to warn doctors and public health officials to implement low-level measures to reduce cardiovascular risk. Prehypertension (PHTN) is such a category. It is a borderline high-risk blood pressure status defined as SBP of 120-139 mmHg and/or DBP of 80-89 mmHg (2). When compared to normotensive patients with a 10% yearly progression rate to hypertension, prehypertension is associated with a fourfold greater chance of developing hypertension (3). Prehypertension is related with a 200 percent increase in cardiovascular disease incidence compared to blood pressure less than 120/80mmHg, in addition to the high chance of conversion to hypertension (2). High blood pressure has becoming more prevalent over the world. In 1975, there were 594 million persons with high blood pressure, according to a pooled study of 1479 research from 174 countries with 19.1 million participants. In 2015, this number grew to 1.13 billion individuals, with the majority of the rise occurring in developing countries. Eighty-eight percent (88%) of deaths caused by high blood pressure now occur in underdeveloped countries (4). Prehypertension was shown to be 38 percent common in a meta-analysis of 242 322 people from 11 countries, including Nigeria (5). The prevalence of prehypertension was determined to be 30.9 percent in a recent meta-analysis of 9 studies completed in Nigeria between 1995 and 2020, involving 16, 241 people (95 percent CI: 22.0-39.7). However, the analysis only included studies from four of the country's six geopolitical zones. In addition, no sex-specific prevalence of prehypertension was reported (6). This study aimed to determine the sexspecific prevalence of prehypertension across the 6 geo-political zones as well as in urban and rural settlements of Nigeria.

Study Area
Nigeria is a western African country with a population of over 250 ethnic groups and a land area of 923,769 square kilometers (7). It is organized into six geopolitical zones or regions, with 36 states and a capital. In 2021, the population is expected to be 211.4 million. About 43.4% of the population was under the age of 14. In addition, 53.9 percent of the population was aged 15 to 64 years old. Only 2.8 percent of the population was above 65 years old (8). Figure 1 shows the map of the country's geopolitical zones.

Inclusion and exclusion criteria
We considered community-based studies conducted on Nigeria's adult population (over the age of 18). SBP of 120-139 mmHg and/or DBP of 80-89 mmHg were used to characterize prehypertension in the studies. We omitted studies involving people under the age of 18, studies involving pregnant women, and all trials conducted in hospitals.

Study search strategy
Google Scholar, PubMed, Global, and African Index Medicus were the online databases used. The terms 'prehypertension,' 'pre-hypertension,' and 'hypertension' were used in the search. For each word containing the word "Nigeria," the search was repeated. The searches took place between March 4 th and March 15 th , 2021. Two reviewers (MAB and MM) independently screened the abstracts and titles of the papers, and studies were selected and excluded based on predetermined criteria. The full-text papers were then screened independently by the same academics in order to choose studies for qualitative and quantitative analyses.

Qualitative Analysis of the Included Studies
A modified version of the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Research Reporting Prevalence Data was used to assess the methodological quality of the included studies (9). The item was decreased from nine to six questions, with each question receiving a score of one for a yes and zero for a no. The maximum possible score is six. If a study received a minimum of five points, it was considered good quality, and if it received less than five points, it was considered poor quality. Two reviewers (AIH and MM) worked independently on the assessment, with AHY resolving any differences. The modified tool used in critical assessment of the included studies is shown in Table 1. According to a recent meta-analysis of studies completed in Nigeria, the minimal sample size for ranking a study as a 'yes' was 323, calculated using the Epitools online calculator (10) and assuming a 30-percent estimated prevalence (6). The quantitative analysis includes only studies that were assessed to have good methodological quality.

Data Extraction and Quantitative Analysis
Two reviewers worked independently to obtain data (MAB and AHY). Prevalence of prehypertension, sample size, settlement (urban/rural), state and location of study, study year, mean age, and sex composition of study participants were all extracted from the studies. For statistical computing, data was loaded into Excel and then imported into the R statistical environment, version 4.1.0 (11). The Metafor Package (12) was used to fit the Random Effects Model for pooling prevalence and the Mixed Effects Model for meta-regression using the inverse variance method with correction of pooled estimate and its variance using Sidik-Jonkman's estimator for between study heterogeneity (13).

-Influence and Heterogeneity Analyses
There were three degrees of inquiry into heterogeneity. Individual studies were initially looked into for influences, which were defined as follows: 1.
The hat value, which is the standardized difference between each study's reported prevalence and the pooled prevalence, was calculated.

2.
Based on rstudent, a normalized distance between each study's projected prevalence and the pooled prevalence.
3. Cook's distance is the distance between the aggregated prevalence when each study is included and when it is excluded.

4.
Based on difference in fits (diffits), which is a distance in standard deviation units between pooled prevalence with the study included and when the research is omitted.
The cutoff values were chosen using the R metafor package. In the generated influence graphs, studies that met these cut-off values were highlighted in red (14). On a second level, gosh analysis, which is also part of the metafor package, was used to look for probable clusters in the studies that were included. On a third level, a metaregression model was fitted using characteristics of the included studies as predictors to see how much of the computed variability can be attributable to study-level factors including sex composition, geopolitical region, settlement, mean age of the subjects, and sample size.
Prediction Intervals built in the meta R package (15) have been reported to help interpret tau2 and I2 as indicators of between-study heterogeneity (16).

-Publication Bias Assessment
The Funnel plot was used to visually evaluate for probable publication bias, which occurred when papers indicating low prevalence were not published and hence were not included in the meta-analysis. Eggar and colleagues (17) used a formal regression technique to test funnel plot asymmetry.

Search Results
The results of the search strategy can be seen in Figure  2.

Methodological features and the Critical Appraisal of the Included Studies
Out of the 63 studies included in the qualitative analysis, only 21 studies were judged to be of good quality and included in the quantitative analysis. Table  2 shows the result of the qualitative analysis.

Characteristics of the Studies Included in the Quantitative Analysis
Twenty-one studies with a total of 25839 participants met the criteria for inclusion and had sufficient methodological quality to be included in the metaanalysis. The research was carried out between 2011 and 2021. There were 7 studies from the South-West region, 6 studies from the South-South, 3 studies from the South-East, 3 studies from the North-Central, one study from the North-West region and one study was conducted in all the six geopolitical regions. There were seven studies in rural areas, thirteen studies in urban areas, and one research that included both rural and urban areas. The average age of the study participants was 40 years in the South-South studies and 47 years in the South-East investigations. The South-West had the most participants in the study (9907), while the North-East had the fewest (1070). Studies from the country's southern areas are often larger in size and contain older individuals (Figures 3  and 4).

Fitting the Meta-analytic Model
Random Effects Model was fitted using the inverse variance method with correction of pooled estimate and its variance using Sidik-Jonkman's estimator for between study heterogeneity. Prevalence was transformed using an arcsine transformation. Figure 5 shows the forest plot of the model.
In Nigeria, the overall prevalence of prehypertension was 34 percent (95% CI: 30-40%). According to the United Nations' most recent statistics (8), Nigeria had an estimated 41.4 million adult prehypertensive persons. The random meta analytic model had a Pvalue of 0.001. The total between-study heterogeneity (tau2) was estimated to be 0.02 (S.E. = 0.005). The percentage of heterogeneity between-studies not explained by sampling error (I2) was 98.6%. The test for heterogeneity revealed Q (df= 24) = 1600, with a P-value of 0.001, showing significant heterogeneity among the studies. The R program meta created a more reasonable prediction interval ranging from 10% to 61 percent. This range shows the prehypertension prevalence rates that future investigations in Nigeria might uncover.

Analysis of Between Study Heterogeneity -Outliers and Influencers
An investigation of likely outliers and influencers was conducted to explain the significant variation. Figure  6 displays the model's radial plot, which reveals no noticeable outliers. The influence analysis plots in Figure 7 corroborate this.

-Gosh Analysis
The probability of clusters within the included studies was investigated using the metafor R package's Gosh analysis. There are no clusters in Figure 8. The included studies did, in fact, form a single cluster with significant heterogeneity.

-Meta-regression using Studies Characteristics as Predictors
The geography of the study was used as the sole predictor, followed by the region, settlement, mean age, and sample size of the included studies. Tau 2 declined from 0.017 in the model without predictors to 0.002 in the model with predictors (a 91% drop). The change in I 2 is also considerable, going from 98.6%, which indicates significant heterogeneity, to 70.3 percent, which indicates moderate heterogeneity. This suggests that differences in study characteristics account for the majority of the heterogeneity between studies. Figure 9 shows the relative changes in the two heterogeneity metrics after fitting the meta-regression models.

Analysis of Publication Bias
The funnel plot of the model is shown in Figure 10. In the plot, there was no noticeable imbalance. The visual assessment of the funnel plot was confirmed by a formal test for plot asymmetry (regression test), which was not statistically significant (P value = 0.079).

Sex-Specific Prevalence of Prehypertension in Nigeria
As seen in Figure 11, males had a higher prehypertension prevalence of 39.1% (95% CI: 30.9%-47.6%) than females, with a prevalence of 28.5% (95% CI: 21.2%-36.4%). The difference was not statistically significant (P = 0.066). Figure 12 shows the prevalence of prehypertension in Nigeria by geopolitical zones. The North-East had the lowest prevalence of prehypertension, at 18%. The North-West, with a prevalence of 43%, had the highest rate of prehypertension. The model's P value is 0.003, showing a statistically significant variation in prehypertension prevalence among the six geopolitical regions.

Difference in Prevalence of Prehypertension between Urban and Rural Settlements in Nigeria
The pooled prevalence in rural settings was 32%, while the pooled prevalence in urban settings was 37%, as shown in Figure 13. The difference (P = 0.307) was not statistically significant.

Discussion
This study's pooled estimate of prehypertension prevalence in Nigeria, 34% (95% CI: 30-40%), is close to but more precise than a recent meta-analysis in the country, which reported a prevalence of 30.9% (95% CI: 22.0-39.7%) (6). The difference in precision between the two meta-analyses can be explained by the difference in the number of studies and individuals included in each; our analysis includes a larger number of studies and individuals, which explains our estimate's higher precision. It's worth noting that prehypertension and hypertension have nearly identical prevalence rates (6). This indicates that twothirds of adult Nigerians have high blood pressure and are at a high risk of acquiring end-stage cardiovascular events such as stroke, coronary heart disease, and chronic kidney disease. The two measures of heterogeneity, I 2 and tau 2 , show different behaviors in this study's analysis of heterogeneity. The former is known to be sensitive to study size, i.e., if the included studies are quite large, the sampling error will be near to zero, and the I2 ratio will approach 100%. (79). The minimal sample size required to measure the prehypertension prevalence of 34% found in this study with a precision of 0.05 and a 95% confidence interval was 345. (10). The papers included in this meta-analysis are relatively large, according to the selection criteria, with an interquartile range of 635 for sample sizes. As a result, even if the between-study heterogeneity is not significant, I 2 may be closer to 100 percent. This is most likely the case because a metaregression with a single predictor (the region where the study was conducted) reduced tau 2 by 35% while leaving I 2 unchanged (Figure 9). Prehypertension is more common in Nigeria than in the United States (80) and Canada (81), equivalent to what was discovered in China (82), Iran (83), and Nepal (84) and lower than in Ghana (85), India (86), and England (81). (Figure 14). In Nigeria, the higher incidence of prehypertension in urban settlements (32% vs. 37%) contrasts with findings from a nationwide study conducted in 2017, which revealed a higher prevalence of prehypertension in rural regions (37.5% vs. 24.7%) (70). The survey's findings are credible under this meta-analytic paradigm, given the overlapping confidence ranges of the pooled prevalence in rural and urban areas (24-39% and 30-44%, respectively). This high prevalence in rural regions may be a result of Nigerian rural areas' urbanization, with the adoption of lifestyles causing an increase in non-communicable diseases (70). Prehypertension was found to be more common in rural regions than in urban areas in Nepal (40.4% vs 29.3%) (23) and India (41.3% vs 40.2%). (86). The three geopolitical regions in the country with the highest prevalence of prehypertension, in descending order, were the North-West (43%), the South-South (36%), and the South-West (33%), with the North-East having the lowest prevalence rate among the six (18%). Only the prevalence estimates in the South-West and North-East had a limited confidence interval, while the estimates in the remaining regions had wide confidence intervals ( Figure 12). The South-Western and South-Southern regions' high rankings are identical to the national poll described above (70).

Conclusion
Prehypertension affects 34% of adult Nigerians (95% confidence interval: 30% -40%), or 41.4 million people. Males had a 39.1% (95% CI: 30.9-47.6%) higher prevalence of prehypertension than females, who had a frequency of 28.5% (95% CI: 21.2-36.4%). The North-East has the lowest prevalence of prehypertension, at 18%. The North-West, with a prevalence of 43%, had the highest rate of prehypertension. Rural areas had a pooled prevalence rate of 32%, whereas urban settlements had a combined prevalence of 37%. There were no statistically significant differences between the sexes or between rural and urban communities. Differences between geopolitical regions, on the other hand, were statistically significant. In conclusion, Nigeria has a significant prevalence of prehypertension, which poses a future risk of hypertension and other cardiovascular problems.