Examining risk factors for hypertension in Ghana: evidence from the Study on Global Ageing and Adult Health

Abstract

Like most countries in sub-Saharan Africa, hypertension contributes substantially to morbidity and mortality in Ghana, yet nationally representative studies that examine the odds of becoming hypertensive among Ghanaians are conspicuously missing. We aimed to fill this void in the literature. The data used for analysis came from the first wave of the Study on Global Ageing and Adult Health (SAGE), collected in Ghana from January 2007 to December 2008 by the World Health Organization (WHO). A total of 5573 respondents were sampled for the study. Random-effects C-log-log models were employed in examining socio-economic, lifestyle and psychosocial factors on the odds of becoming hypertensive in Ghana. Separate models were run for male and females. Results indicated there were strong significant associations between socio-economic, lifestyle and psychosocial factors on the likelihood of becoming hypertensive, among Ghanaian men and women. Compared with the poorest, Ghanaians from wealthy households were significantly more likely to be hypertensive. Educated women, as compared with the uneducated, were also more likely to be hypertensive. Ghanaians who engaged in vigorous or intensive activities continuously, for at least 10 minutes, were significantly less likely to be hypertensive, compared to those who did not. Happier men had lower odds of becoming hypertensive, and depressed women had increased odds of reporting they were hypertensive.

Keywords

Africa Ghana hypertension non-communicable diseases wealth risk factors World Health Organization

Introduction

Globally, non-communicable diseases (NCDs) are the leading causes of morbidity and mortality within and across populations (1). The World Health Organization (WHO) estimated that 36 million people die of NCDs each year, with approximately 80% of such deaths reported in low- and middle-income countries (2). Countries in Africa, south of the Sahara are on the brink of such diseases, given that 35% of deaths in these parts of the world are attributable to NCDs, and this is projected to increase to 65% in 2020 (3,4). With improvement in the economy and rapid urbanization, Ghana’s share of the burden of these diseases continues to increase, and has widely been acknowledged as disturbing, by both researchers and policy makers (5 –7). Hypertension is one such NCD; and is reported to have claimed the lives of many in sub-Saharan Africa, including Ghana. For instance, a press statement released by the Ghana Health Service identified hypertension as the number one killer in Ghana, with almost 70% of all deaths attributed to it (8). Even more deadly is the fact that hypertension is linked to other NCDs such as stroke, diabetes and some cardiovascular diseases.

While the causes of hypertension are generally unknown, several risk factors have been associated with the condition. These risk factors are often explained in terms of how the disease (including other NCDs) has been theorized, especially for countries in sub-Saharan Africa, including Ghana. For instance, in his seminal work on the epidemiological transition in industrialized countries, Abdel Omran (9) argued that structural changes in both the economy and demography of advanced westernized societies marked the completion of the epidemiological transition during which NCDs replaced communicable diseases. Although different from the experiences of western industrialized countries, the epidemiologic transition model, as described by Omran (9), partially explains Ghana’s ongoing health transition. The Ghanaian economy has seen much improvement over the past few years. Gross Domestic Product (GDP) growth was estimated at 13.7% in 2011; the economy was rated the second largest in West Africa, after Nigeria, and one of Africa’s 12 largest (10). More recently, Ghana was classified a lower-middle income country, and the discovery of oil and petroleum have only added to improving its economic prospects.

The economic transformation of the Ghanaian society reflects in her health and mortality profile. For instance, albeit slowly, infant and child mortality continue to decline, while life expectancy continues to increase. The increasing life expectancy has meant a fundamental shift in the age structure of Ghana, especially as people are surviving longer than before. Thus, the prevalence of NCDs such as hypertension may reflect the beginning of an ageing process within the Ghanaian population. It is important to also mention the lifestyle changes that accompany these rapid economic transformations in Ghana. Increasing education, rapid urbanization and improved technology, both at home and the workplace, have led to increased sedentary lifestyles that most studies link with NCDs such as hypertension (11 –13); however, communicable diseases like malaria, cholera and diphtheria, among others, continue to be common. Thus, Ghanaian public policy makers and health professionals are faced with a double burden of disease. In particular, the upsurge of NCDs, including hypertension, poses serious threats to both the economic and health gains achieved in the past. Yet population-based studies that examine NCDs such as hypertension in Ghana are limited (5,7,14).

Bosu (13) corroborates this, as he mentioned that PubMed and Google Scholar database searches of adult hypertension in Ghana between 1970 and 2009 yielded only 15 population-based articles. Also, although useful, the majority of previous studies have been systematic reviews of the literature (13 –15), have only examined the risks of hypertension using data from specific Ghanaian communities (7), and have provided very limited analyses of the odds of becoming hypertensive in Ghana (ones that do not make reference to class, culture and gender differences). This study fills an important research gap, as one of the few, if not the foremost that utilizes nationally representative data to examine risk factors for hypertension in the Ghanaian population. It also adopts what Setel (12) described as ‘a holistic approach’ to understanding such risks, by examining how individual psychosocial, socio-economic, lifestyle and cultural factors influence the odds of becoming hypertensive in Ghana.

Methods

Data

We analysed data from the Study on Global Ageing and Adult Health (SAGE) collected by the WHO between January 2007 and December 2008. The SAGE is an ongoing program that compiles nationally representative longitudinal data on the health and well-being of adult populations aged 50 years and above for six countries: China, Ghana, India, Mexico, Russian Federation and South Africa. SAGE also included a smaller sample of younger adults, aged 18–49 years. For the Ghanaian survey, a stratified multi-stage cluster design was employed to select respondents. The sample was first stratified by administrative region and type of locality, resulting in 20 strata (16). Approximately 235 ‘Enumeration Areas’ were selected as primary sampling units from these localities, out of which 5259 households were surveyed. A total of 5573 respondents (male n = 2799 and female n = 2764) were then sampled from these households. Response rates at both the household and individual levels were 86% and 80%, respectively. Ethical clearance was obtained from the WHO and the local ethical authorities for each participating country, including Ghana.

Measures

The dependent variable used for analysis was self-reported: respondents were asked if they had ever been diagnosed with high blood pressure (hypertension). Independent variables are grouped into socio-economic, psycho-social and lifestyle factors. Socioeconomic predictors included respondents’ education, which was coded (0 = no education, 1 = primary education, 2 = secondary education and 3 = university education); a derived income variable, created from a series of questions tapping the wealth status of households that was coded (0 = poorest, 1 = poorer, 2 = middle, 3 = richer and 4 = richest); and the employment status of participants was coded as (0 = not employed and 1 = employed).

Two psychosocial predictors were included: respondents’ level of happiness was coded as (0 = very happy, 1 = happy, 2 = neither happy nor unhappy, 3 = unhappy and 4 = very unhappy) and whether they had ever been diagnosed with depression was coded as (0 = no and 1 = yes).

Lifestyle variables included a question that asked respondents if they walk or use a bicycle (pedal cycle) for at least 10 minutes, to get to and from places (coded 0 = no and 1 = yes); whether respondents’ work involves vigorous-intensive activity that causes large increases in breathing or heart rate, such as heavy lifting, digging or chopping wood continuously for at least 10 minutes in the last 30 days (coded 0 = no and 1 = yes); and the frequency of consuming drinks that contain alcohol such as beer, wine, spirits, etc. (coded 0 = don’t drink, 1 = not at all frequent, 2 = less than once a month and 3 = more than once a month).

Some socio-demographic and cultural variables were used as controls. These included age of respondents, measured in complete years; ethnicity (0 = Akan, 1 = Ewe, 2 = Ga Adangbe, 3 = Northern Ghanaian ethnic groups and 4 = other ethnic groups); religious denomination (0 = None, 1 = Christian, 2 = Islam, 3 = Traditional and 4 = other); marital status (0 = married/cohabiting, 1 = never married and 2 = divorced/widowed/separated); and place of residence (0 = rural and 1 = urban). Separate models were built for males and females, given the gender patterns in the epidemiology of chronic diseases. Also, we aimed to explore the intersections of gender and the risks of living with hypertension, in this sample.

Analytical strategy

The dependent variable used in this study is dichotomous, but, as shown in Table 1, the cases are unevenly distributed, meaning that use of a probit or logit link function that assumes a symmetrical distribution could produce biased parameter estimates (17,18). As a result, we chose the complementary log-log function, which is better suited for asymmetrical distributions. The standard complementary log-log models are built on the assumption of independence of observations, but SAGE has a hierarchical structure, with participants nested within survey clusters, which could potentially bias the standard errors (SEs). To control for this dependence, we employed random effects models to estimate the magnitude and significance of clustering. The extent of clustering in our models was measured using intra-class correlations. For standard complementary log-log models, this was calculated as the ratio of the variance at the cluster level to the sum of the variances at the individual and cluster levels. That is:

ρ = \frac{σ_{u}^{2}}{σ_{u}^{2} + \frac{π^{2}}{6}}

Table 1.

Distribution of selected dependent and independent variables by gender.

Variables	Male (%)	Female (%)
Ever been diagnosed with hypertension?	n = 2799	n = 2764
No	89.8	85.7^a
Yes	6.20	8.50
Socio-economic variables
Educational background
No education	21.2	36.6^a
Primary education	32.2	34.3
Secondary education	40.1	26.2
Higher education	5.80	2.60
Income quintile
Poorest	13.4	16.9^a
Poorer	17.7	18.7
Middle	19.3	18.7
Richer	21.6	23.2
Richest	27.9	22.5
Employment status
Unemployed	15.0	18.3^a
Employed	85.0	81.7
Lifestyle variables
Walk or bike for 10 min, continuously?
No	18.8	27.1^a
Yes	81.2	72.9
Vigorous-intensive activity for 10 min, continuously?
No	40.8	57.6^a
Yes	59.2	42.4
Frequency of drinking alcohol in the last 12 months?
Don’t drink	37.9	52.5^a
Not at all	8.80	12.0
Less than once a month	14.3	15.4
More than once a month	39.0	20.1
Psycho-social factors
Level of happiness
Unhappy	11.0	8.30^a
Neither	18.5	26.7
Happy	70.5	65.0
Ever been diagnosed with depression?
No	99.1	99.4
Yes	0.90	0.60
Control variables
Average age of respondents	45.12	43.89^a
Marital status
Married	84.3	58.8^a
Never married	8.60	9.10
Divorced/widowed/separated	7.10	32.1
Place of residence
Rural	55.8	52.1^a
Urban	44.2	47.9
Ethnicity
Akan	50.3	55.0^a
Ewe	4.90	5.40
Ga Adangbe	11.3	12.1
Northern ethnic groups	14.0	11.8
Others	19.5	15.7
Religious denomination
None	4.70	2.30^a
Christians	69.0	81.0
Muslims	17.9	10.1
Traditionalists	7.50	5.50
Others	0.90	1.10

p < .01

Here, $σ_{u}^{2}$ is the cluster level variance and $\frac{π^{2}}{6}$ is the variance at Level 1 (individual level), which is that of the standard logistic regression (17,18). The GLLAMM program, available in STATA, was used to build all models.

Results

Table 1 shows the distribution of selected dependent and independent variables by gender, and a description of the sample used for the study. The average age of respondents included in the study sample was approximately 45 years for men and 44 years for women. The majority of respondents were married, lived in rural areas, identified as Akans and were Christians. A higher proportion of Ghanaian women self-reported as hypertensive compared to men. Significant socioeconomic differences were also observed with a higher proportion of men likely to be in wealthy households, employed and highly educated. Also, compared to women, a higher proportion of Ghanaian men were physically active and rated themselves as happier, but consumed more alcohol than their female counterparts.

Bivariate results are shown in Table 2. Socio-economic variables are significant predictors of being hypertensive in Ghana, with highly educated and men and women from wealthy households being more likely to be hypertensive, compared to uneducated and poorer women. However, employed men and women are less likely to be hypertensive, compared with the unemployed.

Table 2.

A bivariate analysis of the odds of becoming hypertensive among the male and female population in Ghana.

Socio-economic variables	MaleOR	FemaleOR
Educational background
No education	1.00	1.00
Primary education	1.59 (0.239)^b	1.60 (0.223)^b
Secondary education	3.42 (0.208)^c	2.54 (0.384)^c
Higher education	5.10 (0.359)^c	9.39 (0.778)^c
Income quintile
Poorest	1.00	1.00
Poorer	1.15 (0.389)	1.68 (0.296)^a
Middle	2.07 (0.356)^b	2.14 (0.310)^c
Richer	5.37 (0.344)^c	4.48 (0.454)^c
Richest	9.58 (0.356)^c	9.03 (0.714)^c
Employment status
Unemployed	1.00	1.00
Employed	0.267 (.282)^c	0.307 (0.378)^c
Lifestyle variables
Walk or bike for 10 min, continuously?
No	1.00	1.00
Yes	0.336 (0.233)^c	0.582 (0.213)^c
Vigorous-intensive activity for 10 min, continuously?
No	1.00	1.00
Yes	0.252 (0.186)^c	0.323 (0.278)^c
Frequency of drinking alcohol in the last 12 months?
Don’t drink	1.00	1.00
Not at all	2.66 (0.283)^c	1.10 (0.211)
Less than once month	2.29 (0.261)^c	1.12 (0.201)
More than once month	0.722 (0.212)	0.415 (0.282)^c
Psycho-social factors
Level of happiness
Unhappy	1.00	1.00
Neither	0.554 (0.328)^a	0.361 (0.252)
Happy	0.364 (0.296)^b	1.03 (0.224)
Ever been diagnosed with depression?
No	1.00	1.00
Yes	1.55 (1.38)	2.77 (0.576)^a
Control variables
Age of respondents	1.028 (0.007)^c	1.037 (0.010)^c
Marital status
Married	1.00	1.00
Never married	0.087 (1.18)^b	0.350 (0.698)
Divorced/widowed/separated	0.845 (0.248)	1.94 (0.270)^c
Place of residence
Rural	1.00	1.00
Urban	4.85 (0.205)^c	4.67 (0.642)^c
Ethnicity
Akan	1.00	1.00
Ewe	1.03 (0.325)	1.36 (0.297)
Ga Adangbe	1.48 (0.248)	1.24 (0.246)
Northern ethnic groups	1.94 (0.230)	1.10 (0.246)
Others	0.858 (0.275)^c	0.426 (0.301)
Religious denomination
None	1.00	1.00
Christians	3.32 (0.474)^c	5.70 (0.666)^c
Muslims	1.41 (0.512)	3.19 (0.658)^a
Traditionalists	0.310 (0.697)	1.11 (0.707)
Others	2.83 (0.869)	3.42 (1.11)

p < .1

p < .05

p < .01

OR are reported and SE are in brackets.

OR: odds ratio; SE: standard error

Lifestyle variables were significantly associated with the odds of becoming hypertensive in Ghana. Men and women who walked or biked continuously for 10 minutes daily were significantly less likely to be hypertensive, compared to those who did not. Similarly, Ghanaians engaged in vigorous or intensive activity continuously for 10 minutes on a daily basis were significantly less likely to be hypertensive. Compared to those who did not consume alcohol at all, men who indicated they consumed alcohol in the last 12 months, and those who consumed alcohol once in less than a month were significantly more likely to be hypertensive. Surprisingly, women who consumed alcohol ‘more than once a month’ were significantly less likely to be hypertensive, compared to those who did not consume alcohol at all. Also, the odds of becoming hypertensive was associated with old age, living in urban areas, being divorced/widowed/separated and identifying as a Christian.

Multivariate results are presented in Table 3 and Table 4. Three separate models were built for male and female participants. Model 1 examined the effects of socio-economic predictors on the likelihood of becoming hypertensive in Ghana, controlling for theoretically relevant demographic variables. Model 2 added lifestyle variables and Model 3 included psychosocial factors. Our results indicated that for Ghanaian men, socio-economic variables were significantly associated with the odds of becoming hypertensive, with Ghanaian men from wealthier households more likely to be hypertensive than poorer men. Ghanaian men who were employed were significantly less likely to be hypertensive, compared to those who were unemployed. When lifestyle factors were controlled, we observed significant attenuation in the odds of Ghanaian men from wealthier households to become hypertensive. The magnitude of the risks for men from wealthier households widened and the coefficients became highly significant when psychosocial factors were controlled. Results show that Ghanaian men who do vigorous work and those identifying as ‘happy’ were significantly less likely to be hypertensive. However, compared to those who not consume alcohol at all, Ghanaian men who did not consume alcohol in the past 12 months and those who consumed alcohol less than once a month were significantly more likely to be hypertensive.

Table 3.

Multivariate analysis of hypertension among men in Ghana, 2008–2009.

Socio-economic variables	Model 1AOR	Model 2AOR	Model 3AOR
Educational background
No education	1.00	1.00	1.00
Primary education	1.29 (0.379)	1.16 (0.268)	1.48 (0.397)
Secondary education	2.06 (0.369)^b	1.56 (0.272)	2.10 (0.389)^a
Higher education	1.64 (0.540)	1.27 (0.355)	1.75 (0.551)
Income quintile
Poorest	1.00	1.00	1.00
Poorer	0.508 (0.570)	0.596 (0.498)	0.554 (0.582)
Middle	1.08 (0.499)	1.06 (0.399)	1.17 (0.517)
Richer	2.75 (0.473)^b	2.10 (0.402)	3.10 (0.503)^b
Richest	3.94(0.488)^c	2.66(0.445)^b	4.62(0.525)^c
Employment status
Unemployed	1.00	1.00	1.00
Employed	0.336(0.306)^c	0.544 (0.318)^a	0.524 (0.310)^b
Lifestyle variables
Walk or bike for 10 min, continuously?
No		1.00	1.00
Yes		0.687 (0.247)	0.607 (0.300)^a
Vigorous-intensive activity for 10 min, continuously
No		1.00	1.00
Yes		0.579(0.247)^b	0.485(0.281)^c
Frequency of drinking alcohol in the last 12 months?
Don’t drink		1.00	1.00
Not at all		1.93 (0.386)	2.43 (0.415)^b
Less than once a month		2.03 (0.395)	2.54 (0.406)^b
More than once a month		0.971 (0.249)	0.933 (0.336)
Psycho-social factors
Level of happiness
Unhappy			1.00
Neither			0.478 (0.421)
Happy			0.248(0.418)^c
Ever been diagnosed with depression?
No			1.00
Yes			1.25 (1.32)
Control variables
Age of respondents	1.05(0.012)^c	1.03(0.014)^b	1.04(0.012)^c
Marital status
Married	1.00	1.00	1.00
Never married	0.125 (1.51)	0.351 (1.11)	0.129 (1.68)
Divorced/widowed/separated	0.607 (0.367)	0.695 (0.268)	0.526 (0.382)
Place of residence
Rural	1.00	1.00	1.00
Urban	3.42(0.280)^c	2.27(0.355)^b	3.03(0.303)^c
Ethnicity
Akan	1.00	1.00	1.00
Ewe	1.07 (0.469)	0.990 (0.318)	0.790 (0.488)
Ga Adangbe	1.37 (0.390)	1.17 (0.246)	1.13 (0.395)
Northern ethnic groups	1.81 (0.385)	1.47 (0.253)	1.60 (0.403)
Others	0.414 (0.441)^b	0.541 (0.334)	0.395 (0.461)^b
Religious denomination
None	1.00	1.00	1.00
Christians	1.91 (0.638)	1.47 (0.454)	1.83 (0.641)
Muslims	1.11 (0.722)	1.26 (0.526)	1.62 (0.755)
Traditionalists	0.442 (0.916)	0.512 (0.651)	0.485 (0.925)
Others	2.42 (1.27)	1.66 (0.876)	2.51 (1.68)
Log-likelihood	−654.131	−642.444	−612.809
Level 2 variance	8.50(3.09)^c	2.59 (3.76)	8.59(3.30)^c

p < .1

p < .05

p < .01

AOR are reported and SE are in brackets.

AOR: adjusted odds ratio; SE: standard error

Table 4.

Multivariate analysis of hypertension among women in Ghana, 2008–2009.

Socio-economic variables	Model 1AOR	Model 2AOR	Model 3AOR
Educational background
No education	1.00	1.00	1.00
Primary education	1.76 (0.234)^c	1.87 (0.241)^c	1.84 (0.247)^c
Secondary education	1.93 (0.297)^b	2.08 (0.350)^c	2.09 (0.295)^c
Higher education	4.18 (0.572)^c	7.54 (0.577)^c	4.06 (0.584)^c
Income quintile
Poorest	1.00	1.00	1.00
Poorer	1.74 (0.325)	1.79 (0.341)	1.75 (0.355)
Middle	2.11 (0.327)^b	2.23 (0.337)^b	2.32 (0.347)^b
Richer	4.39 (0.419)^c	4.81 (0.348)^c	5.42(0.359)^c
Richest	6.36 (0.476)^c	7.77 (0.283)^c	7.85(0.392)^c
Employment status
Unemployed	1.00	1.00	1.00
Employed	0.402 (0.275)^c	0.379 (0.224)^c	0.393 (0.229)^c
Lifestyle variables
Walk or bike for 10 min, continuously?
No		1.00	1.00
Yes		1.10 (0.193)	1.01 (0.202)
Vigorous or intensive activity for 10 min, continuously
No		1.00	1.00
Yes		0.578 (0.219)^c	0.582 (0.299)^c
Frequency of drinking alcohol in the last 12 months?
Don’t drink		1.00	1.00
Not at all		0.884 (0.265)	0.853 (0.275)
Less than once a month		1.02 (0.251)	1.03 (0.258)
More than once a month		0.567 (0.291)^b	0.590 (0.299)
Psycho-social factors
Level of happiness
Unhappy			1.00
Neither			1.06 (0.299)
Happy			1.15 (0.271)
Ever been diagnosed with depression?
No			1.00
Yes			3.13 (0.555)^b
Control variables
Age of respondents	1.03 (0.010)^c	1.03 (0.010)^c	1.03(0.010)^c
Marital status
Married	1.00	1.00	1.00
Never married	0.289 (0.776)	0.214 (0.782)^b	0.221 (0.785)^b
Divorced/widowed/separated	1.68 (0.219)^c	1.76 (0.223)^c	1.79(0.233)^c
Place of residence
Rural	1.00	1.00	1.00
Urban	2.19 (0.239)^c	2.19 (0.206)^c	1.97 (0.210)^c
Ethnicity
Akan	1.00	1.00	1.00
Ewe	1.81 (0.327)^a	1.98 (0.340)^b	1.96 (0.347)^b
Ga Adangbe	1.32 (0.264)	1.41 (0.290)	1.54 (0.298)
Northern ethnic groups	1.25 (0.276)	1.33 (0.3030	1.28 (0.320)
Others	0.891 (0.324)	0.928 (0.357)	0.815 (0.367)
Religious denomination
None	1.00	1.00	1.00
Christians	1.46 (0.715)^b	5.05 (.721)^b	4.66 (0.722)^b
Muslims	3.78 (0.737)^a	3.97 (0.791)^a	4.22 (0.799)^a
Traditionalists	1.98 (0.784)	2.39 (0.852)	2.20 (0.854)
Others	7.31 (1.24)	7.84 (1.28)	8.08 (1.28)
Log-likelihood	−869.342	−863.676	−821.350
Level 2 Variance	3.26 (2.45)	5.06 (1.55)^c	5.12(1.59)^c

p < .1

p < .05

p < .01

AOR are reported and SE are in brackets.

AOR: adjusted odds ratios; SE: standard errors

For Ghanaian women, the effects of socio-economic predictors on the odds of becoming hypertensive was more evident, as both educated women and women from wealthier households had higher odds of becoming hypertensive compared to uneducated and poorer women. Similar to their male counterparts, Ghanaian women who were employed were less likely to be hypertensive, compared to the unemployed. Unlike Ghanaian men from wealthier households whose odds of becoming hypertensive were significantly attenuated by lifestyle factors, the odds for educated women and women from wealthier households were rather accentuated when both lifestyle and psychosocial factors were controlled in Model 2 and Model 3. Women who engaged in vigorous work for at least 10 minutes were less likely to be hypertensive, compared to those who did not.

Those diagnosed with depression were significantly more likely to be hypertensive compared to women without such a diagnosis. For both men and women, being older and living in urban areas were associated with higher odds of becoming hypertensive. While the never married women were less likely to be hypertensive, divorced/separated/widowed women were significantly more likely to be hypertensive compared to married women. Christians were more likely to be hypertensive compared to respondents without any religion. Also, compared to the Akans, Ewe women reported higher risks of hypertension.

Discussion

In his editorial published in the International Journal of Epidemiology, Lloyd-Sherlock (19) asked if hypertension had become the new human immunodeficiency virus (HIV) epidemic. This comparison was mainly because like HIV, hypertension continues to be a major cause of mortality in both rich and poor countries. For instance, Lloyd-Sherlock argued that both HIV and hypertension are asymptomatic, but can lead to fatal illnesses within populations. Also, projections indicate that mortality resulting from hypertension will by far exceed that of HIV/AIDS in the next 20 years (19). Of all the low-income countries, those in sub-Saharan Africa, in particular, have not been spared. Available evidence shows that countries in sub-Saharan Africa contribute substantially to the burden of hypertensive conditions globally and that this is projected to increase by 27% in the next 10 years (19). Yet few comprehensive studies have been conducted examining what predisposes the population to such risks. Using data from Ghana, this study contributes to the literature in this regard and has important practical implications for policy makers.

Overall, our results suggested that Ghanaians with higher socio-economic status (the highly educated and the wealthy) are more likely to be hypertensive. This finding is consistent with others that show that in most sub-Saharan African populations, the risks of hypertension increase with affluence (20,21). While the mechanisms linking socio-economic status (such as education and income) and the risks of hypertension is unclear, some studies point to lifestyle changes that usually accompany the urbanization process. Van der Poel et al. (22) argue, for instance, that economies that are undergoing economic transformations, such as those in sub-Saharan Africa, are equally experiencing major shifts from occupations that are labor intensive and require manual work to those that are capital-intensive and require technology. The transition from a labor-intensive economy to a capital-intensive economy has led to increased sedentary lifestyles, but this may be particularly true for Ghanaians with higher education and skills who may benefit from this transition.

It was not surprising to find that lifestyle factors significantly affected the odds of becoming hypertensive for both men and women in Ghana. For instance, the finding that men and women who do vigorous-intensive activities for more than 10 minutes a day are less likely to be hypertensive is instructive and consistent with other studies that show that moderate-to-vigorous physical activity could offer protection against the risk of becoming hypertensive (20,23 –25). More interesting was the finding that when such lifestyle factors were controlled, the odds of becoming hypertensive reduced significantly for Ghanaian men from wealthy households (see Model 1 and Model 2 of Table 3). This confirms our earlier theoretical stance that the increased sedentary lifestyles that accompany economic success may be what exposes the wealthy and educated to becoming hypertensive, in particular the male respondents. It was also observed that the odds of hypertension are significantly higher for educated women and for Ghanaian women from wealthier households, even more than what was estimated for men, and that this disadvantage was magnified when lifestyle factors were controlled. These results suggested a clear gender difference in the way lifestyle factors confound the relationship between socio-economic predictors and the odds of becoming hypertensive in Ghana. Unemployment appeared to be a major risk factor for increased blood pressure among men and women in Ghana. This finding provides some evidence for the documented advantages that employment could confer on the health of individuals within populations. On the contrary, it indicated how job insecurity could adversely affect health outcomes, including becoming hypertensive (26 –28).

Psychosocial factors had differential effects for both men and women with respect to the odds of becoming hypertensive in Ghana. Compared to the unhappy, Ghanaian men who indicated they were happy were significantly less likely to be hypertensive. More importantly, when such psychosocial factors were controlled, the odds of Ghanaian men from wealthy households becoming hypertensive magnified. Further checking indicated that significantly, the majority of men from wealthy households reported higher levels of depression compared to poorer men; however, for women, level of happiness did not significantly affect their odds of becoming hypertensive, but depression did, as women ever diagnosed with depression were more likely to be hypertensive. These findings demonstrated the long-established links between psychological well-being and physical health (29 –31).

While not the focus of this paper, the significance of some control variables are worth mentioning and discussing. For instance, the finding that urban dwellers were more likely to be hypertensive, compared to rural dwellers was consistent with several other studies that point to urban living as more stressful, and presents lifestyle and dietary changes that increased the risks of contracting NCDs, including hypertension (32 –34). It is intriguing to find that divorced, separated or widowed women were more likely to be hypertensive; and that single women had a lower likelihood of being hypertensive compared to married women. The risks of divorced, separated or widowed women may point to the emotional challenges women face in the event of divorce or the loss of a husband, especially in a society that is patriarchal and has limited social safety nets for women.

The fact that married women had increased odds of becoming hypertensive was inconsistent with evidence elsewhere (35), but may be indicative of Ghanaian women not happy within their matrimonial homes. A cross-classification analysis of marital status and level of happiness indicated the divorced, separated or widowed women were very unhappy, followed by married women. Also, the fact that Ewe women are significantly more likely to be hypertensive compared to the Akans, needs mention. It is known that the Ewes are heavily concentrated in the Volta region, where fishing, salt production and the levels of sodium in their diet may be higher than found in the Akan-speaking areas, where farming is the predominant activity and diets may have less sodium (36).

Despite the interesting findings, several limitations are worth noting. First, the data used here are self-reported, which is often subject to report bias. This is even more problematic, as we know that quite a substantial proportion of Ghanaians live with hypertension, but have not been diagnosed. In this regard, it is possible that the higher reported rates among the wealthy, educated and urban residents may be due to their access to health facilities that enable easy diagnosis of these diseases compared to the poorer, uneducated and rural dwellers. Second, we used only the first wave of the SAGE data and produced cross-sectional estimates, meaning that causal connections cannot be drawn between the dependent and independent variables. Notwithstanding, findings from this study have important policy implications.

The findings suggested that adopting a holistic approach towards dealing with hypertension within the Ghanaian population, one that emphasizes life-style changes, especially among the wealthy, and that improves the psychosocial health of the Ghanaian people in general. It is recommended, for instance, that public education be intensified on physical activity. Our findings also lend support to the WHO’s global recommendation of a national physical activity plan for all member states, including Ghana.

Footnotes

Funding

Research was funded by a grant from the Collaborative Applied Research in Economics (CARE), Department of Economics, Memorial University.

References

World Health Organization (WHO). Report, 2011. Global status report on non-communicable diseases. Geneva: WHO.

World Health Organization (WHO). Non-communicable diseases. Report, 2013. Geneva: WHO.

Amuyunzu-Nyamongo

Need for a multi-factorial, multi-sectorial and multi-disciplinary approach to NCD prevention and control in Africa. Glob Health Promot 2010; 17: S21–32.

Giles

WH.

Preventing non-communicable diseases in sub-Saharan Africa. Glob Health Promot 2010; 17: S3–5.

Aikins

Addo

Offei

Bosu

Agyeman

Ghana’s burden of non-communicable diseases: future directions in research, practice and policy. Ghana Med J 2012; 46: S1–2.

Ministry of Health (MOH). National policy for the prevention and control of chronic non-communicable diseases in Ghana. Accra, Ghana: MOH, 2011.

Addo

Amoah

Koram

KA.

The changing patterns of hypertension in Ghana: a study of four rural communities in the Ga district. Ethnic Dis 2006; 16: 894–899.

Ghana Health Service. Report, 2007. Hypertension is the number one killer disease in Ghana. Accra, Ghana: Ghana Health Service.

Omran

AR.

The epidemiologic transition: a theory of the epidemiology of population change. The Milbank Q 2005; 83: 731–757.

10.

African Development Bank (ADB). Report, 2012. Republic of Ghana country strategy paper, 2012–2016. Tunis-Belvedere, Tunisia: ADB.

11.

Miszkurka

Haddad

Langlois

Freeman

Kouanda

Zunzunegui

MV.

Heavy burden of non-communicable diseases at early age and gender disparities in an adult population of Burkina-Faso: World Health Survey. BMC Pub Health 2012; 12: 24–34.

12.

Setel

PW.

Non-communicable diseases, political economy, and culture in Africa: anthropological applications in emerging pandemic. Ethnic Dis 2003; S149–157.

13.

Bosu

WK.

Epidemic of hypertension in Ghana: a systematic review. BMC Pub Health 2010; 10: 418–432.

14.

Addo

Smeeth

Leon

DA.

Hypertension in sub-Saharan Africa: a systematic review. Hypertension 2007; 50: 1012–1018.

15.

Agyemang

Addo

Bhopal

Aikins

Stronks

Cardiovascular diseases, diabetes and established risk factors among populations of sub-Saharan African descent in Europe: a literature review. Global Health 2009; 5: 7–24.

16.

Biritwum

Report, 2013. Ghana - Study on Global Ageing and Adult Survey, 2007–2008, Wave 1. WHO Multi-country Studies data archive. Geneva: World Health Organization.

17.

Gyimah

Tenkorang

Takyi

Adjei

Fosu

Religion, HIV/AIDS and sexual risk-taking among men in Ghana. J Bio Soc Sci 2010; 42: 531–547.

18.

Tenkorang

Owusu

Correlates of HIV/AIDS testing in Ghana: some evidence from the demographic and health surveys. AIDS Care 2010; 22: 296–307.

19.

Lloyd-Sherlock

Is hypertension the new HIV epidemic?

Int J Epidemiol 2014; 43: 1–3.

20.

Opie

Seedat

YK.

Hypertension in sub-Saharan African populations. Circulation 2005; 112: 3562–3568.

21.

Vijver

Akinyi

Oti

. Status report on hypertension in Africa - consultative review for the 6^th session of the African Union Conference of Health on NCDs. Pan Afr Med J 2013; 16: 38–55.

22.

Van der Poel

O’Donnell

Doorslaer

. Urbanization and the spread of diseases of affluence in China. Report, 2008. HEDG Working Paper, University of York ISSN 1751–1976. York, UK: University of York.

23.

Farrell

Kampart

Kohl

Barlow

Mascera

Paffenbarger

. Influences of cardiorespiratory fitness levels and other predictors on cardiovascular disease mortality in men. Med Sci Sports Exerc 1998; 30: 899–905.

24.

Foy

Foley

Agostino

Goff

Mayer-Davis

Wagenknecht

LE.

Physical activity, insulin sensitivity and hypertension among US adults: findings from the insulin atherosclerosis study. Am J Epidemiol 2006; 10: 921–928.

25.

Mungal-Singh

. Lifestyle changes for hypertension. S Afri Fam Prac 2012; 54: S12–16.

26.

Adler

Newman

. Socio-economic disparities in health: pathways and policies. Health Affair 21: 60–76.

27.

Brackbill

Siegel

Ackerman

SP.

Self-reported hypertension among unemployed people in the United States. Brit Med J 1995; 310–368.

28.

Levenstein

Smith

Kaplan

GA.

Psychosocial predictors of hypertension in men and women. Arch Intern Med 2001; 161: 1341–1346.

29.

Hamer

Batty

Stamatakis

Kivimaki

Hypertension awareness and psychological distress. Hypertension 2010; 56: 547–550.

30.

Grimsrud

Stein

Seedat

Williams

Myer

The association between hypertension and depression and anxiety disorders: results from a nationally-representative sample of South African adults. PLos ONE 2009; 4: e5552.

31.

Stein

Abad

Aguilar-Gaxiola

Alonso

Angermeyer

Demytteneare

. Early childhood adversity and later hypertension: data from the World Mental Health Survey. Ann Clin Psychiatry 2009; 2: 19–28.

32.

Mathenge

Foster

Kuper

Urbanization, ethnicity and cardiovascular risk in a population in transition in Nakuru, Kenya: a population-based study. BMC Pub Health 2010; 10: 569–581.

33.

Peer

Steyn

Lombard

Gwebushe

Levitt

A high burden of hypertension in the urban black population of Cape Town: the Cardiovascular risk in Black South Africans (CRIBSA) study. PLos ONE 2013; 8: e7857.

34.

Seedat

Hackland

DBT

. Prevalence of hypertension in the urban and rural Zulu. J Epidemiol Comm Health 1982; 36: 256–261.

35.

Wang

Effects of marital status and transition on hypertension in Chinese women: a longitudinal study. Paper presented at the 2005 annual meeting of the Population Association of America, 31 March – 2 April 2005, Philadelphia, USA.

36.

Lawson

The structure, migration and resettlement of Ewe fishing units. Afr Stud 1958; 11: 21–27.