Vitamin D status and cardiometabolic disease risk among healthy adults of Northern Ghana

Study Design

A cross-sectional study design was used to investigate the relationship between serum vitamin D status and CMD risk of apparently healthy adults of Northern decent in Ghana.

Study area

The study was conducted in the capital of the Northern Region of Ghana, Tamale among the Dagomba ethnic group. Ghana is located in West Africa, south of the Sahara and on the West of the Gulf of Guinea. The country is divided into 16 administrative regions with the Greater Accra Region as the national and administrative capital. According to the 2021 Population and Housing Census (PHC), the total population of Ghana was 30,832,019 with an annual growth rate of 2.5%. Tamale is the capital town of the Northern Region. It is located within the Guinea Savannah belt and is the fourth largest city in Ghana, with a population of 2, 310, 939 inhabitants who are predominantly Dagombas, and with a growth rate of 7.5% (GSS, 2021 PHC). The size of Tamale is approximately 922km sq. and is considered the fastest growing city in West Africa (http://www.ghanadistricts.com), Tamale is located 600 km north of Accra-the capital city of Ghana. The metropolis experiences one rainy season from April to September or October with a peak in July and August. The mean annual rainfall is 1100 mm within 95 days of rainfall in the form of tropical showers. Consequently, staple crop farming is highly restricted by the short rainy season. (GLSS, 2005/2006). The northern region is one of the poorest of the then three northern regions (now five) of Ghana (GLSS, 2005/2006).The population of adults in Northern Region aged 25 to 59 years according to the latest 2010 population was 741, 318 (GSS, 2010 PHC).

Study Population

The study participants were selected from among free living and apparently healthy individuals aged 25-59 who resided in four selected communities of Tamale. The inclusion criteria were: (1) age> 25 years and < 60 years of age, (2) Ghanaian by ancestral origin and (3) Apparently healthy, defined as no physical complaints at the time of recruitment), no known communicable or non-communicable chronic diseases and cardiovascular disease. Pregnant women or people on lipid lowering drugs/ anti-diabetic drugs/ anti-hypertensive drugs were excluded.

Recruitment and Sampling Technique

Three hundred adult volunteers, aged 25 to 59 years old, who were of northern extraction and Dagomba ancestry, and living in Tamale in the Northern Region were recruited according to the inclusion criteria. The participants were recruited from households in three communities in Tamale, namely Dungu, Dohani and Zujung communities. These three communities were randomly selected from a list of indigenous communities in Tamale. The individual participants were then selected by directly contacting individual adults and tracing them to their homes or going into households to identify adults who were willing to participate in the study. Each participant was informed of the study, and those who were willing to participate were screened for inclusion or otherwise using the screening interview questions. Screening was done by administering the screening interview question to participants. The questions included a set of inclusion and exclusion criteria stated above. Those who met the inclusion criteria were deemed to have qualified for participation. Any participant who passed the screening interview, was asked to observe 12 hours overnight fast for blood samples to be taken. The blood sample of the participants was taken early the next morning and participants completed the main questionnaires.

Socio-demographic and lifestyle data

Data on socio-demographic characteristics and lifestyle were collected using a structured questionnaire. Data collected included alcohol intake, smoking, daily physical exercise and past medical history. For female respondents, contraceptive use and menopause were also elicited. History of chronic diseases and treatments of these chronic diseases, such as hypertension, diabetes mellitus and dyslipidemia were also elicited by the questionnaire.

Assessment of Body Composition

For the purpose of obtaining accurate anthropometric measures, participants were encouraged to empty their bowels and bladder if they had the urge to do so before the anthropometric measures were taken. Also, only light clothing without shoes and cap/hats were accepted as prescribed dressing for the measurement of participants’ weight and height. Participants stood on the stadiometer in an upright posture while their feet were apposed and the height measures were read. The weights of participants were measured using the Body Composition Analyzer (BCA), which automatically calculated the Body Mass Index (BMI) after the height and age parameters of the participants were inputted into the BCA. To measure waist circumference, a tape measure was placed at the midpoint between the lowest rib and the iliac crest, as described by Klein et al., 2007. Three measurements were taken, and means were calculated to the nearest 0.1cm. The measurement of Hip Circumference (HC) was done using the widest point between the iliac crest and buttock as described by Bengtsson et al., (1993). Participants were made to stand in an upright position while their feet were in apposition. The values for the measurements were recorded to the nearest 0.1cm and the mean values calculated from three measurements.

Assessment of blood pressure

The blood pressure (BP) of the participants was also measured. Each participant was made to take 10 minutes of rest before the BP measurement was done. In order to ensure consistency and to obtain reliable and true BP based on recommendation by Pickering et al., 2005, the measurement was done from the left arm of all the participants using a digital sphygmomanometer (OMRON, Japan). Appropriate size cuff was placed on the left upper arm. Both Systolic Blood Pressure (SBP) and Diastolic blood pressure (DBP) were taken in triplicates at five minutes intervals. Blood pressure was categorized as: Normal-SBP < 120mmHg and DBP < 90mmHg. Hypertension-SBP > 140mmHg and DBP > 90mmHg (WHO, 1999).

Dietary assessment

Information on food consumption was obtained using a validated food frequency questionnaire (FFQ). A 3-Day repeated 24hour dietary recall, which included a weekend, was also done to assess dietary intake. Participants were aided to describe all meals and fluid consumed the previous day and to estimate quantities consumed using commonly used household handy measures in Ghana. The quantities of intake were converted to gram equivalents and amounts of nutrients determined from grams of food intake using a template based on Ghanaian foods. Protein, fat, and carbohydrates were calculated as a percentage of total energy intakes.

Biochemical assessments

Blood Biochemical markers measured from blood serum were serum Total Cholesterol (TC), Triglycerides (TG), High-Density Lipoprotein (HDL), Low-Density Lipoprotein (LDL) and Fasting Blood Glucose, as well as glycated haemoglobin (HbA1c). Before the participants’ venous blood samples were taken from the median cubital vein, they were made to fast overnight for twelve hours. The blood samples were taken and handled by phlebotomists who had been given additional training for the purpose of this study. An Automated Biochemistry Analyzer (Hitachi, 7600-020ISE) was used to run the biochemical test at the Medical Laboratory of Tamale Technical University Hospital in Tamale. The tests were carried out by a Principal Biomedical Scientist.

Procedure for Biochemical Analysis

About 7ml of blood sample was collected from via venipuncture from each participant by trained phlebotomist for lipid profile, HbA1c and fasting plasma glucose analysis. Ice chest-containing ice packs were used to temporarily store and transport blood samples from the field to the clinical laboratory of Tamale Technical University for analysis. About 4mls of the blood sample was dispensed into a gel activator tube and the remaining 3mls was dispensed 1.5 ml each into a fluoride tube for fasting plasma glucose (FPG) and an EDTA tube for HbA1c tests. Serum was obtained from the blood samples in the gel activated tube by centrifuging for ten minutes using an eppendorf centrifuge 5804 at 4000 rotations per minute (rpm). Test kits from Medsource Ozone Biomedicals Pvt were used for all the biochemical tests.

Principle of Fasting Plasma Glucose

Glucose oxidase oxidizes glucose present in the plasma to yield gluconic acid and hydrogen peroxide which subsequently reacts with hydroxybenzoate and 4-aminoantipyrine in the presence of peroxide enzyme to give a red-violet quinoneimime dye.

Method

Blood samples in the fluoride tubes were allowed to stand for about 10 minutes. Then the sampled were centrifuged for about 5 minutes. About 10µL of plasma obtained after centrifugation was pipetted into labelled test tubes and 1ml of glucose reagent added. The test tubes were incubated in a water bath at 37oC for about 1o minutes. A blank was then prepared by pipetting about 10µL of deionized water into a labelled test tube and 1ml of glucose reagent added to it. The absorbance of the solution was read at 510nm using a semi-automated spectrophotometer (Biolabo Diagnostic Kenza Biochemistry Try, France) to determine blood glucose concentration.

Glycated Haemoglobin (HbA1c) Test Principle

Haemolyzed whole blood is mixed with weak binding cation-exchanging resin continuously for five (5) minutes. There is adherence to non-glycated haemoglobin to resin during mixing and the glycated haemoglobin is detached in the precipitate. The precipitate binding the glycated haemoglobin is divided from the resin by mixing and filtering. The ratio of absorbance of glycated haemoglobin fraction to total haemoglobin fraction is measured to determine the percentage of glycated haemoglobin (Gabbay et al., 1977; Gonen and Rubenstein, 1978).

Preparation of haemolyte

About 250 µL of lysing reagent was pipetted into labelled sample tubes, then, about 50 µL of properly mixed whole blood was added and incubated at room temperature for complete lysis of erythrocytes to occur.

Preparation of glycated haemoglobin

The toppers of prefilled resin tubes were removed and about 100 µL of haemolysate was added to the resin tubes. The resin separators were inserted in a way that allows the rubber sleeve to be about 1cm above resin suspension surface. The haemolysate in the tubes was mixed for five continuous minutes on a stirrer and afterwards, the resin separators were inserted to firmly pack resin at the bottom for the supernatant to enter the separator tubes. The supernatant was transferred into a cuvette and absorbance was at 415 nm using distilled water as blank.

Preparation of total haemoglobin

An automated pipette was used to measure deionized water into previously labelled test tubes. Then about 20 µL of haemolysate was added to each test tube and properly mixed. Absorbance was read at 415 nm against distilled water blank. Percentage of glycated haemoglobin (GHA1%) was determined using the formular below: Using a conversion table, percentage glycated haemogloin A1c (HbA1c) that corresponds with the % GHbA1 obtained from the calculation above was obtained.

Lipid Profile (Serum Triglyceride-TG, Total Cholesterol-TC, Low Density Lipoprotein-Cholesterol-LDL-C, High Density Lipoprotein Cholesterol-HDL-C)

Gel and clot activator tubes with about 4ml of blood samples were allowed to stand for about 30 minutes after centrifuging for about 10 minutes at 4000rpm to obtain serum. The centrifugation was done using the Eppendorf Centrifuge 5804, EpperdortNetherler-Hinz, Germany. Lipid profile parameters were then determined from the serum using reagents from Forttress Diagnostics Limited, UK.

Serum Triglycerides Test Principle

Note: GPO-glycerol phosphate oxide, GK-glycerol kinase

Method for Determining Serum Triglycerides (TG) Concentration

A sterile automatic micropipette was used to pipette 10µL of serum into a labelled sterile test tube and 1000 µL of triglyceride reagent added. The test tube was gently stirred and incubated for 5 minutes in a 37°C water bath. The concentrations of the triglycerides were determined by reading the absorbent of the red colour dye at 505nm with 1 cm light path cuvette in a semi-automated spectrophotometer.

Total Cholesterol Test Principle

Note: CHE- Cholesterol Estate CHO- Cholesterol Oxidate

Method for Determining Total Cholesterol Concentration

A sterile micropipette was used to pipette about 10µL of serum into a labelled sterile test tube and 1000 µL of cholesterol reagent was added. Another 10 µL of deionized water was put in another test tube and 1000 µL of cholesterol also added to constitute the blank solution. The content of the test tubes were stirred gently and the test tubes were incubated in a 37oC water bath for 5 minutes and then an absorbance was read at 500nm using a semi-automated spectrophotometer.

Determination of Serum HDL-C Concentration

The determination of serum HDL-C was done using a semi-micro assay. With a sterile automatic micropipette, about 200 µL of serum was transferred into a labelled sterile test tube and about 500 µL of HDL precipitant reagent was added. LDL-C became precipitated from the serum leaving thw HDL-C as supernatant after adding phosphotungstic acid and magnesium chloride containing reagents. The test tubes were incubated at room temperature for 10 minutes after gentle mixing, then centrifugation was was done at 4000 rpm for 10 minutes in an Eppendorf Centrifuge. The CHOD-PAP method was then used to estimate HDL-C content. A blank solution was then prepared by pipetting about 100 µL of supernatant serum HDL into sterile test tubes, then 1ml of HDL reagent was added. The test tubes were gently mixed and incubated at 37oC in a water bath for 5 minutes. HDL-C cpncentration was then determined by measuring absorbance at 500 nm using a semi-automated spectrophotometer.

Serum LDL-C Concentration

The LDL cholesterol concentration of all the samples was computed using the Friedewald formular: LDL-C = TC – [HDL-C + (TG/2.2)] mmol/l

Assessment of Serum Vitamin D and Determination of Vitamin D Status

Serum 25-hydroxyvitamin D concentration was determined by electrochemiluminescence immunoassay (Leino et al., 2008) using a Roche Modular E170 Analyzer (Roche Diagnostics GmbH) at the Tamale Teaching Hospital. Participants whose serum 25-hydroxyvitamin D levels were 50nmol/L (20ng/ml)-125nmol/L (50ng/ml); 25 nmol/L (10ng/mL) to 50nmol/L (20ng/ml); and < 25nmol/L (<10ng/ml) were considered as having sufficient; insufficient and deficient vitamin D levels respectively (Arundel, et al., 2012).

Definitions of Anthropometric and Biochemical Parameters

1. Generalized Obesity (BMI) and Central Obesity (WC)

   ➢WC ≥ 102 cm for men and ≥ 88 cm for women

   ➢Overweight: BMI 25.0 – 29.9 kg/m2 Obesity: BMI ≥ 30 kg/m2

   ➢Underweight: BMI < 18.5kg/m2; Normal weight: 18.5 – 24.9kg/m2 (WHO, 2008: WHO, 2006)

   ➢WHR > 0.9 for males and > 0.85 (WHO, 2016)

2. Type 2 Diabetes:

   ➢FPG ≥ 7mmol/l or HbA1c ≥ 6.5% (WHO 2006)

3. Serum Lipid Profile

   ➢Elevated triglycerides ≥ 1.7mmol/l;

   ➢Elevated total cholesterol ≥ 6.5mmol/l

   ➢Elevated LDL ≥ 4.1

   ➢Decreased HDL < 1.03mmol/l for men and < 1.3 for females

   (NCEP ATP III, 2001)

4. Blood Pressure

   ➢Normal Blood Pressure: SBP 120 – 140 mmHg

   ➢High Blood Pressure: SBP > 140 mmHg and DBP > 90mmHg (WHO, 1999)

5. Coronary Risk

   ➢High Coronary Risk: TC/HDL-C ≥ 3.5

   ➢Low Coronary Risk: TC/HDL-C < 3.5

Diagnosis of Cardiometabolic Disease (CMD)

The diagnosis of CMD in this study was in accordance with the criteria provided by the WHO (WHO 2007). Central obesity was defined according to the WHO criteria: WC ≥ 102 cm for men and ≥ 88 cm for women or Waist-to-Hip Ratio (WHR) ≥ 0.90 in men and ≥ 0.85 in women and a WHTR of > 0.50; Plus, any two of the following; raised triglycerides (>1.7mmol/L); reduced HDL-cholesterol (<1.03mmol/L in men and <1.29 mmol/L in women); raised blood pressure (systolic >140 mmHg or diastolic >90 mmHg); raised fasting plasma glucose (≥ 5.6 mmol/L).

Ethical Consideration

This study was approved by the Council for Scientific and Industrial Research (CSIR) Institutional Review Board (IRB) (Ref: RPN 003/CSIR-IRB/2018). In addition, this study was approved by the Metropolitan Director of Health Services, Tamale (TMHD/MPHs/09). All participants signed informed consent prior to their participation.

Statistical Analysis

Data was entered into Microsoft excel and analyzed using Statistical Package for Social Sciences (SPSS), IBM Corporation, USA, version 25. The variables were tested for normality by the Kolmogorov-Smirnov Test. Socio-demographic characteristics, anthropometric and biochemical parameters were analyzed using descriptive statistics and expressed as percentages, frequencies, means and standard deviations. Means of continuous variables were compared using one-way analysis of variants (ANOVA). Correlation analysis was performed to determine associations between metabolic factors and vitamin D. The data were expressed as mean ± SD or median (inter-quartile range). T-test was performed to compare two groups of data with normal distribution, while Mann-Whitney U test was performed to compare two or more groups of the data with non-normal distribution. Binary regression analysis was used for evaluating the effects of the variables of CMD on the serum vitamin D status. Multivariate logistic regression analysis was used for evaluating the effects of vitamin D and other risk factors of Cardiometabolic Disease (CMD). A p-value of less than 0.050 was considered as statistically significant.

Prevalence of CMD risk factors

The mean age of the 302 participants was 38.78 (±10.42) years, similar between the two genders. Table 1 presents summary of risk factors for CMD. The females recorded significantly (p=0.006) higher mean BMI (24.31± 7.96) than the males (22.52± 3.07), as well as total % body fat (males=17.60%±7.90%, females=32.09%±9.50%, p<0.001), waist (males=80.74± 9.80, females=85.00±12.27, p=0.002) and hip circumferences (males=93.45, ±9.34, females=96.89±12.10, p=0.011). On the other hand the males recorded higher serum lipid parameters, such as mean serum total cholesterol (5.74± 1.02) and LDL (3.57 ±1.02) than the females (TC=5.40±1.05, p=0.008; LDL=3.22±1.09, p=0.005). The mean ages, SBP, DBP, FBG, HbA1c, triglycerides, HDL and visceral fat were similar between males and females.

Cardiometabolic Disease Prevalence

Figure 1 shows the prevalence of CMD, determined by being centrally obese, plus two CMD risk factors. Of the total population, 6.4% had CMD, all women. Of the female population alone, 10.2% had CMD.

• Download figure
• Open in new tab

Fig 1:

Prevalence of CMD among Study Participants.

Vitamin D levels and status by age and gender of participants

The distribution of serum vitamin D status is shown in Table 2 below. Seven percent (7%) of the participants had vitamin D deficiency (<25 nmol/L), 7.5% in the females and 6.5% among the males (p=0.349). The mean (±SD) vitamin D levels of those with vitamin D deficiency was 20.1 ± 2.4 nmol/L; that of vitamin D insufficiency was 41.2 ± 7.2 nmol/L and vitamin D sufficiency was 80.2 ±17.1 nmol/L. The mean (±SD) vitamin D for the study group was 65.2 nmol/L, 66.6 (± 25.3 for males and 63.8 (± 25.8) for females, with no statistical significance between the two categories (p=0.349). Mean serum vitamin D levels for the different vitamin D categories by age group were different (p=0.047). Vitamin D deficiency increased with age, from 5% in the 20-29 years category to 12% among the 50-59 year olds, while vitamin D sufficiency decreased with age, from 74.4% in the 20-29 year olds to 48% among the 50-59 years category.

Relationship between Serum Vitamin D and CMD Risk Factors

Table 3 compares the mean CMD risk factors between participants with vitamin D deficiency, insufficiency and sufficiency. Apart from serum total cholesterol, HDL and triglycerides, all the CMD risk factors were different between the different vitamin D status categories. Mean age of those with VDD is significantly higher than those with VDS (p=0.009). Both mean SBP (P<0.001) and DBP (P<0.001) increased with VDD, with the least in the VDS, followed by VDI and the highest in the VDD. Consistently, the means for the following: FBG (p=0.032), BMI (p=0.004), HbA1c (p=0.040), and LDL (p=0.047) are highest in the VDD group, followed by the VDI group and lowest in the VDS group. Therefore, all the CMD risk factors seem to increase with vitamin D deficiency. Similarly, all the body composition parameters, being mean visceral fat (p=0.003), percent total body fat p<0.001), WC (p=0.007) and WHR (p=0.002) are all significantly highest in the VDD group and lowest in the VDS group, with the exception of HC. CMD also increased with vitamin D deficiency. Close to 30% of participants with vitamin D deficiency had CMD compared to 5.85 in the vitamin D insufficiency and 4.1% in the vitamin sufficiency group (p<0.001). Overweight and obesity (using BMI cutoffs) were also higher among the vitamin D deficiency group (47.6%) and lowest in the vitamin sufficiency group (18.46%) (p<0.001). Similarly, more than 4 in 10 participants with central obesity were vitamin D deficient, compared with about 3 in 10 in the vitamin D insufficiency group and less than 2 in 10 in the vitamin D sufficient group.

Binary Regression Analysis to Predict Vitamin D and C Risk

A binary regression (logistic regression) was performed to ascertain the effects of Age, Gender, SBP, DBP, FPG, HbA1c, STG, SC, LDL, HDL, BF, WC, HC, WHR, Visceral fat, and BMI on the likelihood that a participant or respondent’s CMD risk factor will have an influence on vitamin D deficiency. The logistic regression model was statistically significant, χ² =40.589, p <0.001. The model explained 36.4% (Nagelkerke R Square) of the variance in Vitamin D deficiency levels and correctly classified 91.4% of cases. From the results in Table 4, SBP (p = 0.003) and FBS (p = 0.002) added significantly to the model/prediction. The information in the “Variables in the Equation” table can be used to predict the probability of an event occurring based on a one-unit change in an independent variable when all other independent variables are kept constant. As such the table shows that a unit increase in SBP and FPG were associated with a decrease likelihood of exhibiting Vitamin D deficiency levels. The odds of having Vitamin D deficiency are 0.055 (5.5%) times greater in participants with high SBP whilst the odds of having Vitamin D deficiency are 0.076 (7.6%) times greater in participants with high FPG. The results from the binary regression analysis showed that participants with normal SBP had reduced odds (OR 0.055, 95 CI) for vitamin D deficiency compared with those with high SBP. Also, those with low FBS had reduced odds (0.076, 95 CI) for vitamin D deficiency compared with those with high FBS or diabetics. For LDL, those with normal LDL levels had reduced odds for vitamin D (OR 0.362 95 CI) compared with those with high LDL levels.

Predictors of Cardiometabolic Disease (CMD)

From multivariate analysis, filtering FBS, Serum TG, LDL and WC filtered out by in-and-out stepwise method showed FPG, Serum Triglycerides, LDL, and WC as independent risk factors of CMD with p-values of 0.020, 0.013, 0.001 and 0.001 respectively.

The dependent variable (CMD), was categorized into two; those who have CMD were assigned 1, and those who do not assigned 0. As such the dependent variable (CMD) is dichotomous (binary). Logistic regression is the appropriate regression analysis to conduct when the dependent variable is dichotomous (binary). Logistic regression is used to describe data to explain the relationship between one dependent binary variable and one or more nominal, ordinal interval or ratic level independent variables. With regards to this study, we are interested in how the independent variables: BP, WC, Body fat, visceral fat, Serum TG, FPG, LDL, and obesity have influenced the probability of having CMD. In applying the logistic regression, the main assumption of no multicollinearity, no outliers and dependent variables been dichotomous was checked.