Assessment of Association between Antihypertensive Drug Use and Occurrence of New-onset Diabetes in South Indian Patients.

OBJECTIVE: The objective of this study was to determine the association between antihypertensive drug use and new-onset diabetes (NOD) in patients with hypertension (HTN).
MATERIALS AND METHODS: A retrospective observational study was conducted in a tertiary care hospital located in south India. Patients diagnosed with HTN and HTN with diabetes between January 2012 and December 2016, were identified and studied. Medical records of these patients from Medical Record Department were evaluated by medical record review method, and relevant data were recorded in a case record form. Statistical evaluation by chi-square method and odds ratio (OR) was carried out to appraise the incidence of NOD in patients taking antihypertensive medications.
RESULTS: A total of 1250 patients with HTN were screened, and based on inclusion and exclusion criteria, 952 patients were enrolled in the study; among which, 537 were patients with HTN and 415 were patients with diabetic HTN. The majority of the patients with HTN and patients with diabetic HTN were from the age-group of above 60 years. The most commonly prescribed drugs observed in our study population were amlodipine in 94 (22.7%) patients. OR was calculated and it was observed that the combination therapy has a fivefold risk for the development of NOD in patients with HTN, followed by Angiotensin II receptor blockers (2.06) (confidence interval [CI]: 1.56-2.73), diuretics (1.33) (CI: 0.95-1.85), non-Dihydropyridine calcium channel blockers (DHP CCBs) (1.3) (CI: 0.51-3.30), vasodilators (1.13) (CI: 0.40-3.15), and Angiotensin converting enzyme inhibitors (1.06) (0.68-1.64).
CONCLUSION: Patients on non-DHP CCBs, diuretics, and combination antihypertensives showed more chances of developing NOD. Copyright:

INTRODUCTION

Hypertension (HTN) is a long-term medical condition in which the blood pressure (BP) in the arteries is persistently elevated. Recent studies indicate that HTN is present in 25%–30% urban and 10%–20% rural subjects in India. The Global Burden of Disease study has reported that HTN is the most important cardiovascular risk factor in this region and responsible for the largest burden of disease and mortality. Present estimates suggest that a 2mm Hg population-wide decrease in systolic BP (SBP) can lead to prevention of more than 151,000 stroke and 153,000 coronary heart disease deaths in India.[1]

Diabetes and its complications cause a major problem to upcoming public health resources worldwide. On the basis of the collection of data from various parts of the world, World Health Organization (WHO) has proposed that the highest rise in diabetes would happen in India.[234] Socioeconomic, behavioral, nutritional, and public health concerns have caused increase in type 2 diabetes, obesity, and cardiovascular disease (CVD) in several countries.[25] Due to the common pathogenic mechanisms underlying diabetes and CVD, a mutual rise is often seen in the prevalence of both the diseases.[2] Patients with type 2 diabetes, metabolic syndrome X, and prediabetes are often treated with antihypertensive medications because of concomitant HTN or CVD.

Patients with HTN have greater prevalence of insulin resistance, thereby risk for diabetes, compared to normotensive individuals. Post hoc analyses of HTN clinical trial data show that therapy with both diuretics and beta-blockers enhances the risk of development of new-onset diabetes (NOD) suggesting thiazides and beta-blockers in patients with HTN are at greater risk of developing diabetes.[678] It is not clear whether NOD reflects a separate entity associated with greater risk or merely reflects accelerated presentation of diabetes mellitus (DM) in individuals destined to develop DM.[9]

Selection of antihypertensive therapy is a critical problem as most of the classes of antihypertensive medications vary due to their metabolic effects.[10] NOD is common among patients with HTN and it results in CVD. NOD can be reduced by monitoring hypokalemia with a potassium-sparing diuretic and/or potassium supplementation, or by including a potassium-conserving antihypertensive drug such as an Angiotensin converting enzyme inhibitors (ACEI), Angiotensin II receptor blocker (ARB), or an anti-aldosterone agent.[11] Previous research has showed that beta-blockers and thiazides may aid the occurrence of type 2 DM.[12] The rationale behind this research was to determine the relationship between antihypertensive drug use and subsequent development of NOD in south Indian patients.

The objective of this study was to assess the association between antihypertensive drug use and occurrence of NOD in south Indian patients.

MATERIALS AND METHODS

The study was a retrospective observational study conducted in the tertiary care hospital of south India from September 2017 to March 2018 (eight months). The study was approved by the institutional ethics committee with IEC No.: 563/2017.

Study population

The study group consisted of patients with HTN (ICD code: I10) and patients with diabetic HTN (ICD code: E11.9) admitted in the general medicine ward during 2012–2016, who fulfilled the following study criteria:

- Inclusion criteria: all patients with HTN above 18 years, and patients with diabetic HTN who developed NOD after 180 days of initiating antihypertensive medication.

- Exclusion criteria: gestational diabetes, type 1 DM, steroid-induced diabetes, and patients who developed diabetes before diagnosis of HTN.

Sample size

The sample size of this study was calculated by relative precision method. From previous literature, incidence of NOD in patients with HTN has been found to be 12%. To estimate 12% incidence of NOD among subjects with HTN with 15% relative precision and 95% confidence interval (CI), the total sample size required is 1250 (250 patients with HTN from each year from 2012 to 2016). A total of 952 were eligible to be included in the study. In these 952 patients, 537 were patients with HTN and 415 were patients with diabetic HTN.

Sources of data

Patient case records, which are kept at the Medical Records Department, include patient demographic information, physician order sheet, drug prescription charts, and laboratory parameters.

Study materials

Case record form (CRF): Demographic information such as patient identification number, gender, age, diagnostic codes, current smoking status, familial history of diabetes, body mass index (BMI), BP, total cholesterol (TC), triglyceride (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL), fasting blood sugar (FBS), serum creatinine (S. Cr), and drugs prescription information was estimated and recorded in the CRF.

Operational modality

Patients diagnosed with HTN and HTN with diabetes in the tertiary care hospital between January 2012 and December 2016, were identified and studied. Medical records of the patients from hospital files and hospital data system were evaluated by medical record review method and recorded in the CRF. Selected patients were further assessed to check if they met the requirements for inclusion and exclusion criteria.

The total sample size required for the study was calculated and found to be 1250 patients (i.e., 250 patients per year from 2012 to 2016). A total of 250 patients of each year were divided into two subgroups: 125 patients with HTN and 125 patients with diabetic HTN. Using MS Excel, 250 random numbers for each year were selected. These random numbers were the serial numbers of the patients to be taken for the study. This randomization method was repeated for each year and their results were manually checked for duplicates. The inpatient numbers of patients, which corresponded to the random serial numbers were collected, and the list for sample size was prepared. This process of randomization minimized the sample bias in our study.

Statistical analysis was carried out using IBM SPSS version 20.0 (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY: IBM Corp.), to appraise the incidence of NOD in patients taking antihypertensive medications, by analyzing the following parameters:

Demographic characteristics such as age, gender and BMI, social habits, occupation, length of hospital stay, duration of disease, complications and comorbidities of disease, past medication history, and drug usage pattern.

Laboratory parameters such as BP, TC, TG, LDL, HDL, fasting blood glucose, and S. Cr were analyzed. Glycosylated hemoglobin (HbA1c) was used to identify diabetes or glucose intolerance. According to hospital standard, HbA1c was classified as normal (4.5%–7%) and high (>7%). BMI was calculated from the available height and weight of the patients. BMI was categorized into underweight (<18.5kg/m2), normal (18.5–24.9kg/m2), overweight (25–29kg/m2), and obese (30–40kg/m2) according to WHO classification.

Statistical analysis

Categorical and continuous variables were summarized using frequency and percentage in descriptive statistics. To calculate the risk of development of NOD in patients with HTN, Pearson chi-square test was applied for different classes of antihypertensive drugs, and odds ratio (OR) was calculated to identify the risk of antihypertensive drugs causing diabetes using SPSS, version 20.0.

RESULTS

A total of 1250 patients with HTN were selected according to the inclusion and exclusion criteria, of which 952 were eligible to be included in the study. In these 952 patients, 537 were with HTN and 415 were with diabetic HTN.

Demography of the patients

Gender-wise distribution

Among the 537 patients with HTN, 319 (59.4%) were male patients and 218 (40.6%) were females. In the diabetic HTN group, 236 (56.9%) were male patients and 179 (43.1%) were female patients.

Age-group-wise distribution

Of 537 patients with HTN, 291 (52.5%) were older than 60 years of age, 43 (8%) were between 40 and 60 years of age, and 2 (0.37%) were in the age-group of 20–40 years.

In 415 patients with diabetic HTN, 239 (57.6%) were older than 60 years of age, 160 (38.6%) were between 40 and 60 years of age, and 16 (3.9%) fell in the age range of 20–40 years.

Distribution of smokers

In 537 patients with HTN, 77.3% of the patients were nonsmokers, 13% were smokers, and 8% were reformed smokers. Among the patients with diabetic HTN, 86.3% were found to be nonsmokers, 8% were smokers, and the remaining 5.5% were reformed smokers.

Distribution of alcohol consumption

Among 537 patients with HTN, 80.1% were found to be nonalcoholics, 15.1% were alcoholics, and 3.4% of them were reformed alcoholics. In 415 patients with diabetic HTN, 82.6% were nonalcoholics, 14.3% were alcoholics, and 3.1% were reformed alcoholics.

Distribution of body mass index

In 537 patients with HTN, 224 (40.4%) had a normal BMI range, whereas 118 (21.3%) of patients were overweight, 53 (9.6%) were obese, and 44 (7.9%) were underweight. Among 415 patients with diabetic HTN, majority of the patients (155 [37.3%]) had a normal BMI, with the rest of them being overweight (122 [29.4%]), obese (62 [14.9%]), and underweight (4 [1%]), which is more clearly shown in Figures 1 and 2.

Figure 1

Distribution of body mass index of hypertensive patients

Figure 2

Distribution of body mass index of diabetic hypertensive patients

Gender distribution among different age-group of patients with diabetic hypertension

On evaluation of gender distribution among different age-groups in patients with diabetic HTN, it was observed that the majority of the patients were found in the age-group >60 years with 127 (53%) males and 112 (47%) females, followed by the age-group of 40–60 years with 97 (60%) males and 63 (40%) females. The least number of patients was observed in the age-group of 20–40 years, with 12 (75%) males and 4 (25%) females. A better representation can be seen in Figure 3.

Figure 3

Gender distribution among different age-groups of patients with diabetic hypertension

Distribution of complications of hypertension

Of 537 patients with HTN in this study, only 253 had complications. In that, the most common complications observed were ischemic heart disease (IHD) in 65 cases (12.1%), chronic kidney disease (CKD) in 55 cases (10.2%), angina in 52 cases (9.7%), cerebrovascular accident (CVA) in 23 cases (4.3%), atrial fibrillation and heart failure in nine cases (1.7%), and dyslipidemia and angina + IHD in eight cases (1.5%). Detailed representation can be seen in Figure 4.

Figure 4

Distribution of complications of hypertension

Distribution of complications of patients with diabetic hypertension

In the 415 patients with diabetic HTN, only 193 had complications. The most common complications observed were IHD in 48 (11.6%) cases, CKD in 44 (10.6%) cases and angina in 26 (6.3%) cases, foot ulcer in 14 (3.4%) cases, CVA and diabetic retinopathy in 9 (2.2%) cases, and coronary artery disease in 8 (1.9%) cases. Detailed representation can be found in Figure 5.

Figure 5

Distribution of complications of patients with diabetic hypertension

Baseline demographic characteristics of patients with hypertension and diabetic hypertension

The demographic parameters of both patients with HTN and diabetic HTN were compared by taking the mean and standard deviation for age, SBP, diastolic blood pressure (DBP), TC, TGs, LDL, HDL, fasting blood glucose (FBS), random blood sugar (RBS), postprandial blood glucose (PPBS), HbA1c, and S. Cr, and by taking the frequency and percentage for BMI, gender (male, female), and distribution of smokers and alcoholics [Table 1].

Table 1

Baseline characteristics of all patients

Parameter	Hypertension + diabetes mellitus (N = 415)	Hypertension (N = 537)
Age (years)	61.9 ± 10.8	60.2 ± 13.5
BMI ≥ 25 (%)	122 (29.4)	118 (21.3)
Male (%)	236 (56.9)	319 (57.6)
Female (%)	179 (43.1)	218 (39.4)
Smokers (%)	33 (8)	70 (12.6)
Alcoholics (%)	59 (14.2)	81 (14.6)
SBP (mm Hg)	141.9 ± 22.5	144.1 ± 23.9
DBP (mm Hg)	86.3 ± 10.5	87.9 ± 12.7
TC (mg/dL)	169.8 ± 50.1	180.4 ± 52.6
TG (mg/dL)	153.5 ± 83.8	131.2 ± 83
LDL (mg/dL)	99.8 ± 42.1	115.6 ± 71.6
HDL (mg/dL)	38.8 ± 15.1	43.8 ± 16.9
FBS (mg/dL)	176.1 ± 71.6	115.8 ± 38.5
RBS (mg/dL)	210.6 ± 104.1	121.6 ± 43.1
PPBS (mg/dL)	235 ± 90.8	139.9 ± 40.7
HbA1c (%)	8.28 ± 2.0	6 ± 0.83
S. Cr (%)	1.71 ± 5.67	1.4 ± 1.9

History of antihypertensive drugs in patients with diabetic hypertension

It was observed that the most commonly prescribed drugs taken by patients with diabetic HTN were amlodipine, amlodipine + atenolol, telmisartan, losartan, and metoprolol. Detailed history of antihypertensive drugs is represented in Table 2.

Table 2

Past medication history of patients with diabetic hypertension

Antihypertensive drug/s taken by patient before occurrence of diabetes	Number of patients (%) (N = 415)
Amlodipine	94 (22.7)
Losartan	24 (5.8)
ACEI	21 (5.1)
Cilnidipine	14 (3.4)
Metoprolol	25 (6)
Telmisartan	29 (7)
Losartan + thiazide	16 (3.9)
Amlodipine + atenolol	30 (7.2)
Telmisartan +DHP CCB	25 (6)
Telmisartan + thiazide	12 (2.9)
Others	125 (30)

Prescription pattern of patients with hypertension and hypertension with diabetes

General drug usage pattern in patients with hypertension and diabetic hypertension

In patients with HTN, the most frequently prescribed drug classes were Dihydropyridine calcium channel blockers (DHP CCBs) in 352 (63.5%) patients, followed by beta-blockers in 162 (29.2%), ARBs in 128 (23.1%), and diuretics in 87 (15.7%) patients. Combinations of two or more drugs in 358 (64.6%) patients were observed in majority of the patients.

In patients with diabetic HTN, the most frequently prescribed drug classes were DHP CCBs in 227 (54.7%), ARBs in 163 (39.3%), beta-blockers in 104 (25.1%), and diuretics in 85 (20.5%) patients. Combinations of two or more drugs were observed in majority of the patients, that is, 381 (91.8%) patients [Table 3].

Table 3

Prescription pattern of patients with hypertension and diabetic hypertension

Type of antihypertensive drug	Number of patients (%)
	Hypertension patients (N = 537)	Hypertension + diabetes mellitus patients (N = 415)
Diuretics	87 (15.7)	85 (20.5)
Beta-blockers	162 (29.2)	104 (25.1)
DHP CCB	352 (63.5)	227 (54.7)
Non-DHP CCB	9 (1.6)	9 (2.2)
Alpha-blockers	52 (9.4)	32 (7.7)
ACEI	49 (8.8)	40 (9.6)
ARB	128 (23.1)	163 (39.3)
Vasodilators	8 (1.4)	7 (1.7)
Combinations	358 (64.6)	381 (91.8)

Diuretic drug usage pattern in patients with hypertension and diabetic hypertension

Of the 88 hypertensive patients who took diuretics, 48 (8.9%) patients had loop diuretic and 29 (5.4%) had thiazide diuretic and Of the 85 diabetic hypertensive patients who were prescribed diuretics, 51 (12.3%) patients had loop diuretic and 23 (5.5%) had thiazide diuretic as shown in Table 4.

Table 4

Types of diuretics used by patients with hypertension and diabetic hypertension

Type of diuretic	Number (%)
	Hypertension (N = 537)	Hypertension + diabetes mellitus (N = 415)
Thiazide	29 (5.4)	23 (5.5)
Loop diuretic	48 (8.9)	51 (12.3)
Potassium sparing	-	1 (0.2)
Aldosterone antagonist	3 (0.56)	-
Loop + aldosterone antagonist	7 (1.3)	6 (1.4)
Loop + potassium sparing	1(0.18)	-
Thiazide + potassium sparing	-	1 (0.2)
Loop + aldosterone antagonist	-	6 (1.4)
Thiazide + loop	-	2 (0.5)
Thiazide + loop + aldosterone	-	1 (0.2)
antagonist
Total	88 (16.4)	85 (20.5)

Odds ratio for antihypertensive drugs prescribed for diabetic hypertensive drugs

When OR was calculated for various classes of antihypertensive drugs prescribed for patients with diabetic HTN, it was found that for ARBs (2.06), diuretics (1.33), non-DHP CCBs (1.3), vasodilators (1.13), and ACEI (1.06) had a significant chance to develop NOD due to exposure to these drugs. Combinations of drugs showed an OR of 5.60 as compared to monotherapy of drug [Table 5].

Table 5

Odds ratio for different antihypertensive drug classes prescribed in patients with diabetes

Antihypertensive drugs	Adjusted odds ratio	Adjusted 95% confidence interval
Diuretics	1.33	0.957–1.85
Beta-blockers	0.77	0.58–1.03
DHP CCBs	0.63	0.48–0.82
Non-DHP CCBs	1.3	0.51–3.30
Alpha-blockers	0.77	0.49–1.23
ACEI	1.06	0.68–1.64
ARB	2.06	1.56–2.73
Vasodilators	1.13	0.40–3.15
Combination	5.60	3.77–8.30

DISCUSSION

Due to the limited available studies, to identify the association between antihypertensive drug therapy and impaired sugar levels in patients with HTN in Indian setup, a retrospective study was planned out in a tertiary hospital in south India. A total of 1250 patients with HTN, during five years (2012–2016) were screened and based on inclusion and exclusion criteria, 952 patients were enrolled in the study among which, 537 were patients with HTN and 415 were patients with diabetic HTN.

Of 537 patients with HTN in this study, 319 (59.4%) were male and 218 (40.6%) were female patients. In the diabetic hypertensive group, 236 (56.9%) were male and 179 (43.1%) were female patients. It was found that more male patients were admitted for treatment when compared to female patients. This observation was consistent with the study conducted by Ahmad et al.[13] and Niklason et al.[14]

In this study population, mean age for subjects with HTN was 60.2 ± 13.5 years and that of subjects with diabetic HTN was 61.9 ± 10.8 years. Majority of the patients with HTN and diabetic HTN were from the age-group of above 60 years. Age is a risk factor in patients with HTN, which contributes to type 2 DM. A similar study conducted by Ahmad et al.[13] suggested that HTN was found to be mostly observed in patients with the age-group of 51–60 years (46.3%) and subsequently in the age-group of 41–50 (32.2%) years. Another study by Berraho et al.[15] showed that in 525 patients with diabetic HTN, 47.1% of patients were >60 years. HTN among type 2 diabetics was associated[16171819] with age; this association agrees with research literature and with the findings of other studies.

It was observed in our study that most of the patients with HTN (77.3%) and diabetic HTN (86.3%) were nonsmokers. This is compatible with the result of a study carried out by Ahmed and Musallam,[20] in which most of the study participants (90.7%) were not smokers. Our results also matched with that of Berraho et al.[15] who reported that there was no significant association between HTN and smoking. This finding is in contrary with the findings reported in the research literature.[21]

Majority of the patients with HTN and diabetic HTN in our study were not alcoholics. However, Lip et al.[9] did a study, which stated that alcohol users had a lower risk of developing NOD.

BMI is considered to be one of the risk factors for the development of diabetes. We observed that in our study population, percentage of overweight and obese patients (BMI > 25) was found to be high in patients with diabetic HTN when compared with patients with HTN. Dale et al.[22] noted that patients with diabetic HTN had a higher percentage of BMI ≥ 25, especially among women. Chinedu and Nicholas[23] found a direct positive relationship between body weight or BMI and BP,[24] and over 75% of newly diagnosed patients with type 2 diabetes were obese.[25]

Obesity is one of the important factors, which has direct association with diabetes. Prevalence of obesity in this study was found to be more in subjects with diabetic HTN (69, 14.2%), particularly in women when compared to their hypertensive counterparts (53, 9.6%). This finding is similar to the study undertaken by Regensteiner et al.,[26] in which the prevalence of obesity increased from 32% to 38% in women >60 years of age. Subjects receiving multidrug antihypertensive had a higher BMI and more often had coexisting diseases than those taking no medication.[12]

The most commonly prescribed drugs observed in our study population were amlodipine in 94 (22.7%) patients, amlodipine + atenolol in 30 (7.2%) patients, telmisartan in 29 (7%), losartan in 24 (5.8%), and metoprolol in 25 (6%), which is consistent with another Indian study that had reported amlodipine was the most commonly prescribed individual drug. Atenolol was the commonly prescribed drug among beta-blockers, followed by metoprolol, carvedilol, and nebivolol.[13]

Even though current guidelines suggest using ACEI and ARBs as first-line drugs for the treatment of patients of HTN with diabetes, in our study, it was amlodipine, which was prescribed the most, both as monotherapy and in combination with other drugs, which is consistent with the findings of a similar study conducted by Acharya et al.[27] As for patients with HTN, diuretics are considered as the drug of choice for treatment, but in our study, we found that amlodipine was prescribed more. In our study, we calculated OR to identify the risk of antihypertensive drugs causing diabetes. We observed that combination therapy has a fivefold risk for the development of NOD in patients with HTN, followed by ARBs (2.06) (CI: 1.56–2.73), diuretics (1.33) (CI: 0.95–1.85), non-DHP CCBs (1.3) (CI: 0.51–3.30), vasodilators (1.13) (CI: 0.40–3.15), and ACEI (1.06) (0.68–1.64). In diuretics, loop diuretics with 51 (12.3%) users and thiazide diuretics with 23 (5.5%) users were used in most patients. Commonly prescribed combinations include amlodipine–atenolol, amlodipine–telmisartan, and thiazide–telmisartan; probably, the cumulative effect of the drug classes has a higher risk to cause diabetes in patients with HTN.

A study conducted by Huang et al.[28] observed that thiazide diuretics (OR, 1.65; 95% CI: 1.12–2.45) and non-DHP CCBs (OR, 1.96; 95% CI: 1.01–3.75) had a significantly greater influence in causing NOD in patients with HTN. However, beta-blockers, DHP CCBs, alpha-blockers, ACEI, ARBs, and vasodilators were not associated with increased risk of NOD.

Strengths and limitations: A large sample size of 1250 patients, during a span of five years (2012–2016), has been collected and analyzed in this project and an extensive literature review has been carried out. This study provides a systematic overview of available data and it contributes to identify knowledge gaps based on the data from previous studies conducted globally. To the best of our knowledge, this is the only study of its kind that has been conducted in south India. More similar studies should be encouraged to get a clear idea about the incidence of NOD in the Indian population.

Since it was a retrospective study, the limitation of this study was incomplete documentation of relevant information in medical records of the patients.

CONCLUSION

Patients on non-DHP CCBs, diuretics, and combination antihypertensives showed more chances of developing NOD. These data might aid the physician while prescribing antihypertensive drugs to patients having other risk factors for diabetes. It is recommended to carry out more similar prospective studies in future because such studies are required to reveal the risk of NOD among patients with HTN treated with antihypertensive drugs.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.