Analysis of glycemic status in diabetes-naïve patients on statins: A hospital-based cross-sectional study.

INTRODUCTION: Randomized controlled trials, observational studies, and meta-analysis suggest risk of hyperglycemia in patients on statins, and this association is being viewed with renewed interest globally. The present study has tried to explore the possible diabetogenic effect of statins, the mechanism of this effect, and various comorbidities associated with this causation.
MATERIALS AND METHODS: This cross-sectional study was carried out at the Department of Cardiology from October 2015 to March 2017. Patients on statins for at least 1 year and normoglycemic at the time of statin initiation were recruited in the study. The outcome of the present study was development of new-onset diabetes mellitus (NODM). Blood glucose levels and insulin levels were estimated. Other adverse reactions of statins and associated comorbidities in the patients were recorded. Descriptive statistics were used to analyze adverse drug reactions.
RESULTS: A total of 104 patients met the inclusion criteria, of which eight patients (7.7%) developed NODM and 4 (3.8%) developed prediabetes. Atorvastatin 40 mg was most commonly prescribed statin. About 25% of patients taking atorvastatin 80 mg developed diabetes.
CONCLUSION: Statins have a mild-to-moderate risk of developing NODM. The dose of statins is an important factor that increases the risk of diabetes in statin users.

Introduction

With the turn of the century, there has been a sudden shift in the disease burden from communicable diseases to noncommunicable diseases. According to the Global Action Plan 2013 released by WHO, by 2020 non-communicable diseases will account for 80% of the global disease burden, causing a seven out of ten deaths in the developing country and half of them being premature deaths below the age of 70.[1] Premature mortality regarding life lost due to cardiovascular diseases (CVDs) in India has increased drastically from 23.2 million in 1990 to over 37 million in 2010.

The results of the Framingham Heart study in 1948 revealed a correlation between hypercholesterolemia and CVD.[2] Following up from that study, the researchers found that 3-hydroxy-3-methylglutaryl-CoA reductase is a natural target and rate-limiting enzyme involved in the cholesterol biosynthetic pathway.[3] Hence, statins have proven to be the major lifeline in the management of CAD patients. Given the ever-burgeoning menace of CAD on a global scale, an observational study conducted by IMS Health from 2006 to 2010 revealed the monthly statin prescription increased from 45.8 to 84.1/1000 patients with coronary heart disease (CHD).[4] It was shown that Indians have the lowest high-density lipoprotein-cholesterol (HDL-C) and recent studies have revealed that a low HDL-C and high TC/HDL ratio is one of the major predictor for CAD and Indians have a similar pattern of lipid profile.[5] Therefore, intensive statin therapy is recommended in the South Asian population. The recent association of statins with development of new onset diabetes (NOD) has been of utmost interest to the researchers.[6]

Increase in the risk of NOD by disrupting the insulin signaling pathways and by reducing the insulin secretion and or systemic insulin sensitivity has been reported in many studies.[7] With this background, the present study was undertaken to analyze the glycemic status of patients on statins.

Materials and Methods

It was a prospective observational study carried out for a period of 1½ year (October 2015–March 2017) in the Department of Cardiology, Kasturba Medical College, Manipal, Karnataka. The Institutional Ethics Committee approval was taken before the start of the study and written informed consent was obtained from all the participants before their inclusion in the study.

Patients of either sex, above 30 years of age and on statins for at least 1 year with a fasting blood glucose level <100 mg/dl at the time of initiation of statin were included in the study. Diabetics, patients on any concomitant medications that may raise blood glucose levels such as fluoroquinolones, beta-blockers, atypical antipsychotics, glucocorticoids, thiazide diuretics, protease inhibitors, and pregnant women and lactating women were excluded from the study. Patient's detailed history focusing on risk factors if any for metabolic syndrome and family history was taken. Demographic, clinical characteristics and drug history were recorded as per the pro forma.

The primary outcome of the study was the development of NOD mellitus (NODM) and development of prediabetes. Blood glucose levels were estimated in all patients included in the study. In all the patients who had developed diabetes, their blood samples were sent for fasting blood glucose and insulin levels. Possible mechanism of the development of diabetes was elucidated using a computerized model homeostatic model assessment (HOMA) which is a method to quantify the insulin resistance and beta-cell function and is calculated using the following equation:[8]

Quantitative insulin sensitivity check index (QUICKI) was obtained by the following formula:

1/(log (fasting insulin μU/mL) + log (fasting glucose mg/dL).[9]

In addition, side effects with statins were also recorded.

Data analysis was performed using SPSS version 23 (IBM Corp., Armonk, NY) and Excel software (Microsoft Windows Operating system, version 2016. Licensce: Trialware). For baseline and demographic characteristics, the descriptive analysis was used. Continuous variables are expressed as mean ± standard deviation if normally distributed. Categorical variables were expressed as number of patients and percentages (n, %). Skewed data were expressed as a median and interquartile range (quartile 1 and quartile 3). Risk factors were analyzed as frequencies and percentages. Adverse drug reactions (ADRs) were analyzed using descriptive statistics. Survival analysis and Cox regression model for hazards ratio could not be done in the present study as the sample size was too small to make a definite conclusion regarding the same. Hence, descriptive statistics had to be followed.

Results

A total of 104 patients were enrolled into the study as per the inclusion criteria. Data regarding patient demographics and clinical information were collected in a prestructured pro forma. Baseline demographic parameters have been expressed as the number of patients (n) and the percentage as shown in Table 1. Concomitant medications taken by patients are shown in Table 1. The study population consisted of 66 (63.5%) males and 38 (36.5%) females. Majority of the patients were in the age group of 60–65 years.

Table 1

Baseline demographic parameters of patients included in the study

	Number of patients (%)
Patient characteristics
Male/female	66/38 (63.5/36.5)
Mean age (years)	62.26±10.02
Hypertension	101 (97.2)
Hypercholesterolemia	102 (98.2)
Overweight/obese	35 (36.6)
Ischemic heart disease	70 (67.3)
Concomitant medications
Antiplatelet drugs: Aspirin	93 (89.4)
Clopidogrel	32 (30.76)
Diuretics (except thiazide diuretics)	16 (15.38)
Antianginal drugs	12 (11.53)
ACEIs/ARBs	95 (91.34)
Others	9 (8.65)

Values are expressed as number of patients and percentage. ARBs=Angiotensin receptor blockers, ACEIs=Angiotensin-converting enzyme inhibitors

Among all the 104 patients who received statin therapy, 8 (7.7%) patients developed diabetes mellitus (DM) and 4 (3.8%) patients developed prediabetes.

Figure 1 shows the percentage use of various statins in the present study. Atorvastatin 40 mg and 80 mg were the most frequently and least frequently used statins, respectively. Further analysis showed that of the eight patients on atorvastatin 80 mg, two developed diabetes (25% incidence). Similarly, the chance of development of NODM with atorvastatin 20 mg and 40 mg was 4.1% and14.7%, respectively, and none of the patients on rosuvastatin 20 mg developed diabetes or prediabetes. Body mass index (BMI) of 59 patients (56.7%) was in the normal category, 29 (27.9%) in the overweight category, and 8 (7.7%) each in both obese and underweight category.

Figure 1

Percentage use of different types of statins in the study population

Only one point lipid profile was taken and the frequency distribution of lipid profile among the study population is shown in Table 2.

Table 2

Mean distribution of lipid profile in the study population

Parameters	Mean±SD
Triglycerides (mg/dl)	137±40.83
HDL (mg/dl)	37.11±7.67
LDL (mg/dl)	128.72±42.18

Values represent mean±SD. HDL=High-density lipoprotein, LDL=Low-density lipoprotein, SD=Standard deviation

Percentage of smokers was 52.6%, alcoholics 57.9%, first-degree relative as diabetic were 63.2%, and family history of cardiovascular disorder was seen in 47.4%. Percentage of patients with mild depression and various other psychiatric conditions was 15.4%. Physical inactivity was seen in 68.4% of patients and 31.6% of patients did mild-to-moderate exercises. The percentage of patients with various conditions suggestive of insulin resistance was 21.1% and these included polycystic ovary diseases, acanthosis nigricans. The mean waist circumference among our study population was 86.41 ± 9.13 cm.

Analysis of side effect profile among the study population

Out of 104 patients, 33 (31.73%) patients developed ADRs. Statin-induced myalgia was the most common ADR observed among the population (48.4%). The various side effects observed in the study population are shown in Table 3. ADRs associated with statin were classified according to the WHO-UMC causality assessment scale and have been shown in Table 3.

Table 3

Side-effect profile of patients on statins

ADRs	Number of adverse reactions	Out of total 33 adverse effects (%)	Causality assessment
Myalgia	16	48.4	Possible
Tingling sensation	7	21.21	Possible
Headache	6	18.18	Possible
Loss of appetite	2	6.06	Possible
Dizziness	1	3.03	Possible
Hepatitis	1	3.03	Probable

Values are expressed as number of adverse reactions and percentage. ADRs=Adverse drug reactions

Analysis of outcome of statin use on insulin resistance by homeostatic model assessment and quantitative insulin sensitivity check index

HOMA value of >2.27 is considered as insulin resistance. A total of 10 (83.3%) patients who developed diabetes had a HOMA score of >2.27 which could indicate that a probable mechanism of the development of diabetes among statin users could be insulin resistance. QUICKI score of <0.357 is considered as severe insulin resistance, 10 (83.3%) patients had severe insulin resistance.

Discussion

CVDs are one of the leading cause of deaths in certain parts of the world and in India.[5] Data emerging from clinical trials prompted the US Food and Drug Administration to add a safety label to statins indicating an increase in glycated hemoglobin and fasting glucose levels.[10] Several meta-analyses have proven that the association of statins with adverse glycemic control is real, though its mechanistic confirmation is still mired with controversy. The study conducted by the Canadian Network for Drug Observational studies revealed that the incidence of diabetes was highest in the first 4 months after starting of statin therapy.[11] However, the famous West of Scotland Coronary Prevention Study (WOSCOPS) trial reported a 30% reduction of diabetes with pravastatin 40 mg.[12]

Statins, the most effective drugs in hypercholesterolemia, have an important role in primary and secondary prevention of cardiovascular disorders. There has been a rising concern about the use of statins and development of diabetes in recent years. As statins are a commonly prescribed drug and diabetes affects a large population worldwide, the magnitude of the association between statin use and risk of diabetes needs to be elucidated. In this context, the present study was undertaken to evaluate the effect of statins on the glycemic status of nondiabetic patients on statins for at least 1 year. Furthermore, this study aims at exploring the role of insulin sensitivity in statin-induced diabetes.

The JUPITER (Justification for the Use of Statins in Prevention: An Intervention Trial Evaluating Rosuvastatin) trial revealed that NOD was seen in 27% subjects in the rosuvastatin group compared to the placebo group.[13] Study conducted by Sattar et al. demonstrated a 9% increased risk of incident diabetes among non diabetic statin users.[14]

Demographic parameters such as older age, increased weight, and metabolic syndrome can predict the development of diabetes and is also associated with a threefold risk of atherosclerosis and a 12-fold risk of DM. Therefore, the use of statins can unmask DM in patients with other risk factors.[15]

The mean age of the study population was 62.26 ± 10.02 years. Similar results were reported in the Anglo-Scandinavian Cardiac Outcomes Trial – Lipid-Lowering Arm (ASCOT-LLA) trial where the mean age of the development of diabetes was 63 years and the study conducted by Aiman et al.[16] where the mean age of the patients developing diabetes was 66 years. Studies conducted by Sattar et al. also confirmed that the age group between 55 and 76 years was more at risk for the development of diabetes.[14] Preiss et al. and Navarese et al. also found that the age group of 55–65 years was more at risk of developing diabetes.[1718] Male preponderance was seen among patients on statins (63.5%) in the present study. The ASCOT-LLA trial concluded that about 81% of participants included in their trial were males, in the Long-Term Intervention With Pravastatin In Ischemic Disease (LIPID) trial, the male population was 80% and JUPITER trial had 61.8% of males.[192021] Normal BMI was found in 56.7% of patients in the present study whereas 27.9% were overweight and 7.7% were in the obese and underweight category. This is in accordance with the study conducted by Cassuto et al., who reported that obesity and statin therapy are independently associated with an enhanced risk for the development of diabetes.[22] However, in WOSCOPS, LIPID, HPS, and ASCOTT trials, larger proportions of patients were in the overweight category. This could be the probable reason for higher proportions of diabetes in these trials in comparison to the present study. A study conducted by Laakso and Kuusisto has put forth that BMI increased by 1.3 among statin users and only by 0.4 in the nonuser group (P < 0.02). These new findings indicate that the increase in BMI was faster among the statin users compared to nonstatin users and over time, calorie and fat intake increased in people taking statins.[23] The primary outcome of our study was designed to be the development of NODM. 7.7% (n = 8) of patients developed NODM and 3.8% (n = 4) developed prediabetes within a median time interval of 4 years. The most common statin that was associated with NODM was the higher-potency atorvastatin 80 mg. PROSPER trial which compared pravastatin 40 mg to control, a 6.6% incidence of NODM was observed among the statin group within a median follow-up of 3.2 years. The diabetogenic potential of statin was shown to be dose-dependent.

Meta-analysis conducted by Preiss et al. in 2011 also confirmed that higher potency statins are associated with a 12% risk of NODM compared to lower potency statins over a median follow-up of 4.9 years (odds ratio 1.12, 1.04–1.22).[17]

Patients with major risk factors are more prone to developing NODM. A meta-analysis conducted by Navarese et al. proposed that the incidence of DM with atorvastatin 80 mg, rosuvastatin 20 mg, and pravastatin 40 mg was 15%, 25%, and 7% respectively.[18] Five population-based studies have reported this incidence to be between 18% and 99%.[2425]

Among the various statins prescribed to our patients, atorvastatin 40 mg was the most commonly prescribed statin. About 58.6% (n = 61) were on atorvastatin 40 mg, 23% (n = 24) were on atorvastatin 20 mg, 7.6% (n = 8) on atorvastatin 80 mg, and 10.5% (n = 11) on rosuvastatin 20 mg. However, the appearance of NODM and prediabetes was more likely seen with atorvastatin 80 mg (i.e., 25%) compared to atorvastatin 40 mg (14.7%) and atorvastatin 20 mg (4.1%). This is in accordance with a study conducted by Koh et al. who put forth that atorvastatin 80 mg was associated with higher incidence of diabetes in the study and a dose-dependent effect was seen among the study population.[26] The ASCOT-LLA, LIPID, and PROSPER trials have also concluded that the incidence of diabetes is more with higher doses such as atorvastatin 80 mg than with lower doses such as atorvastatin 10 mg, 20 mg, and lower-potency rosuvastatin 5 mg.[2728] In the present study, only two statins, that is, atorvastatin and rosuvastatin were prescribed. Drug utilization studies from different parts of India also show that atorvastatin is most frequently used statin followed by rosuvastatin and simvastatin.[2930] Studies mentioning the use of other statins are scarce.

Increase risk of diabetes (46%) within a median follow-up of 5.9 years was noted in the Metabolic Syndrome in Men study among patients on atorvastatin 40–80 mg and simvastatin 10–20 mg. The study was conducted on 8749 nondiabetic men, to detect the insulin sensitivity. About 24% reduction in insulin sensitivity and 12% reduction of insulin secretion (P < 0.01) were noted in the statin-treated arm.[29] A Korean study has revealed a significant increase in the development of NODM among statin users following a median period of 3 years.[30]

The associated comorbidities in our study population were hypertension and hyperlipidemia which was seen in 97.2% and 98.2% of patients, respectively. This was consistent with the Collaborative Atorvastatin Diabetes Study (CARDS) trial where an 84% incidence of hypertension was recorded among the atorvastatin arm compared to the placebo arm[31] The Framingham heart study also reveals that individuals with a family history of CVD have a higher risk of developing obesity, hyperglycemia, high blood pressure or dyslipidemia. Interactions between individual behavior, genetic factors and environment could also be hypothesized as a cause for CVD.[32]

Insulin resistance is defined as a normal or elevated insulin level producing an attenuated biological response. It is related to various medical conditions such as fatty liver, atherosclerosis, obesity, skin tags, acanthosis nigricans, and polycystic ovary diseases. The percentages of patients with various conditions suggestive of insulin resistance were 21.1%. Many large epidemiologic studies have confirmed measuring waist circumference (WC) as a strong correlate of clinical outcome. The American Diabetes Association has confirmed that the measurement of WC provides incremental value in predicting diabetes, CHD, and mortality rate above and beyond that provided by BMI.[33] The mean waist circumference of the study population was 86.41 ± 9.13 cm.

Thirty-three ADRs were reported in the study group. Of these, myalgia was the most commonly reported (48.4%) followed by tingling sensation (21.21%). Statin-induced hepatitis was the only ADR which required discontinuation of the drug. Myopathy and rhabdomyolysis were not observed in the present study. Causality assessment of adverse events was done using Naranjo's ADR Probability Scale.[34] These scales assessed whether the adverse effects were definitely, probably, or possibly associated with the drug or were doubtful. All of the observed adverse reactions showed possible causality to statin use. Regarding severity, all the adverse events were mild, being of levels 1 and 2 by the Hartwig Severity Assessment Scale.[35] There was a single case of statin-induced hepatitis of probable causality and drug was withdrawn due to the same.

A HOMA score of >2.27 is considered as insulin resistance and 83.3% of the population of patients in the study population who developed diabetes had a HOMA score of >2.27. QUICKI score of <0.357 is considered as severe insulin resistance, and 83.3% of patients had a severe insulin resistance in the present study. This could indicate that the mechanism of statin-induced diabetes could be insulin resistance. This proposed mechanism can be one of the multiple mechanisms by which statins increase the risk of diabetes.

The strength of the study is that there have been only few studies from India which evaluated the association of statins with NODM in Indian population. In the present study, the association of statin-induced NODM has been seen in South Indian population. This study has also tried to elucidate the possible mechanism of developing hyperglycemia in this patient population. Further, various factors which can increase the risk of NODM in patients on statins have also been studied.

The major limitation of our study was the sample size. The sample size was calculated considering incidence to be 6% based on the previous studies. However, a survival analysis could not be done as the sample size was not adequately powered to detect a statistical significance with the risk factors. Being a cross-sectional study, it shows an association between statin and NODM and not causation since exposure and outcome are measured at the same time point.

Conclusion

The role of statins in the prevention of cardiovascular disorders is well established. However, there is a need to watch for worsening glycemia in anticipation when intensifying the statin treatment. Regular monitoring of blood glucose levels is thus warranted in patients on statins such as high doses of atorvastatin and more so in patients with risk factors predisposing to DM. In addition, further studies with control group can confirm the effect of statins on glycemic control.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.