Optimal risk thresholds for prescribing statins as primary prevention of cardiovascular disease in Iranian general population: a benefit-harm modelling study

Abstract

Purpose

The use of statins for the primary prevention of cardiovascular diseases (CVD) is associated with various beneficial outcomes, alongside certain undesirable effects. This study aims to determine optimal risk thresholds above which statin therapy yields a net benefit, considering both the positive effects and potential adverse effects, as well as their probabilities and patient preferences.

Methods

Quantitative benefit-harm balance modeling was applied to the Iranian general population aged 40 to 75 years with no history of CVD. The analysis utilized data from prior studies, including statin effect estimates for different outcomes from a meta-analysis, patient preferences obtained from an Iranian survey, and baseline incidence rates of adverse outcomes sourced from the Global Burden of Disease study for Iran. Outcomes were defined as angina, myocardial infarction, fatal coronary heart disease, fatal or non-fatal stroke, and heart failure. Benefit-harm balance indices were calculated for various combinations of age, sex, and 10-year CVD risk.

Results

Statin therapy was found to be advantageous at a lower 10-year CVD risk threshold in men (18–23%) compared to women (24–28%). Furthermore, individuals aged 40–45 years exhibited a lower risk threshold (18% in men, 24% in women) than those aged 70–75 years (23% in men, 28% in women).

Conclusion

The desirable 10-year risk thresholds for statin prescription in the primary prevention of CVD vary by age and gender, ranging from 18 to 28%, encompassing a spectrum of outcomes from angina to CVD mortality. These results suggest hard-CVD risk thresholds of 7.5% to 10% for both sexes.

Introduction

Cardiovascular disease (CVD) is a major global public health problem and a leading cause of mortality, responsible for more than 30% of deaths worldwide and 46% in Iran, annually [1]. Elevated levels of total cholesterol and low-density lipoprotein (LDL) are established risk factors for CVD, making cholesterol management a critical goal in CVD prevention [2]. Statins, a class of lipid-lowering drugs, are commonly prescribed for the primary prevention of CVD in high-risk individuals, according to national and international guidelines [3]. Although statins have been shown to have significant benefits, their use is not without risks. Several studies have reported different benefits and harms associated with statin use [4], and the appropriateness of prescribing statins is influenced by age, sex and population [5]. Therefore, there is a need to determine an optimal risk threshold for prescribing statins that balances their potential benefits and harms in a given population [6].

The American College of Cardiology (ACC) and the American Heart Association (AHA) issued a clinical guideline in 2013 that lowered the CVD risk threshold for initiating lipid-lowering therapy to 5–7.5%, which was criticized for over-prescribing statins to healthy individuals [7–11]. In 2018, the ACC/AHA guideline was revised to recommend statin use for individuals with a 10-year risk of 5–20%, taking physician and patient preferences into account [12] and in 2022, the United States Preventive Services Taskforce (USPSTF) recommended initiating statin therapy in those with a 10-year risk of ≥ 10% and at least one traditional CVD risk factor [13]. The 2021 European Society of Cardiology (ESC) has categorized people as very high CVD risk when risk is ≥ 7.5% for age under 50, ≥ 10% for age 50–69, and ≥ 15% for age > 70 [14]. World Health Organization (WHO), in the HEARTS protocols, recommends statin therapy for CVD risks ≥ 20%, the risk threshold which is considered for statin therapy in primary healthcare centers in Iran [15].

This study aims to evaluate the benefit-harm balance of statin therapy for primary prevention of CVD in an Iranian population and identify the 10-year risk threshold by age and sex that maximizes the net benefit of statins while minimizing potential harms.

Methods

One of the most well-known approaches to assess the benefit-harm balance of health interventions with both favourable and adverse consequences is the Gail/NCI method [16]. This method, which was proposed by Dr. Michael Gill et al. at the National Cancer Institute of the United States, calculates the benefit-harm balance index (i.e., net benefit) by taking into account the different consequences (both benefits and harms) attributed to the intervention, the baseline risk of these consequences (outcomes), and the significance of these outcomes [17, 18]. In the forthcoming sections, more detail of the statistical analysis of this modelling approach adopted in this study will be presented.

Description of information

The Gail/NCI methodology was applied to the information from prior studies, including statin effect estimates from a meta-analysis and data provided by Henock G. Yebyo et al. [4, 5], patient preferences from an Iranian survey performed for the current study [19], and baseline incidence of harm outcomes and competing risk for non-CVD death from the global burden of disease study for Iranian population [20]. The present study models the benefit–harm balance of statins for people aged 40 to 75 years with no history of CVD. Cardiovascular outcomes were defined as total CVD events including moderate MI, severe MI, moderate stroke, severe stroke, unstable angina and heart failure as used in our previous study to find people’s preferences and inform the current study [19]. Benefit–harm balance indices were calculated for different categories of sex, age group, and 10-year CVD risk.

The necessary parameters for assessing the benefit-harm balance of statin are presented in more details as follows:

• Beneficial and harmful consequences of statin therapy

The current study focuses on the low to moderate doses of statins (e.g., 10–20 mg for atorvastatin and 5–10 mg for rosuvastatin) typically prescribed for the primary prevention of CVD [21]. We considered different consequences of statin therapy in the benefit-harm balance index concerning their magnitude and significance. These outcomes include favourable outcomes (i.e. the outcomes which are prevented by statins) including fatal and non-fatal outcomes of CVD (including moderate to severe heart attacks, moderate to severe stroke, heart failure and angina) and adverse outcomes including liver and kidney dysfunctions, myalgia, cancer, type 2 diabetes, nausea and headache, cataracts, hemorrhagic stroke and other side effects (depression, sleep disorders) which were identified in the literature review [4].

Information on the effects of statins on the study outcomes is presented in Supplementary Table 1, based on previous studies in the field [5, 6].

• Baseline outcome risks in the population

To estimate the absolute number of consequences that increase or decrease due to statin intervention, it is essential to consider the baseline outcome risks in the population. For this purpose, information on the basic risk of each studied outcome was extracted from various sources. Specifically, data on type 2 diabetes, cancers, hemorrhagic stroke, and total non-CVD deaths were obtained from the Global Burden of Disease (GBD) for Iran [20], while information about myalgia, cataracts, and kidney and liver dysfunction was extracted from observational studies [22]. Since data on the consequences of nausea/headache and discontinuation of treatment due to side effects was unavailable, rates reported in the control group of randomized controlled trials were used, similar to a previous study [23]. Table 1 presents the basic risk level of the studied outcomes considered in calculating the benefit–harm balance index.

Table 1.

Baseline risks (annual rates) of statin-related outcomes from Global Burden of Disease project (Iran data) or other related studies

Outcome	Sex	Age	Rate	95% Confidence Interval
Type 2 diabetes (19)	Men	40 to 44	0.00512	0.00372 - 0.00664
	Men	45 to 49	0.00633	0.00436 - 0.00892
	Men	50 to 54	0.00741	0.00588 - 0.00972
	Men	55 to 59	0.00834	0.00588 - 0.01127
	Men	60 to 64	0.00859	0.00637 - 0.0118
	Men	65 to 69	0.00833	0.00582 - 0.01101
	Men	70 to 75	0.00772	0.00531 - 0.01027
	Women	40 to 44	0.00666	0.00501 - 0.00842
	Women	45 to 49	0.00853	0.00597 - 0.01176
	Women	50 to 54	0.00983	0.00749 - 0.01271
	Women	55 to 59	0.01060	0.00777 - 0.01404
	Women	60 to 64	0.00990	0.00732 - 0.01303
	Women	65 to 69	0.00828	0.00581 - 0.01082
	Women	70 to 75	0.00708	0.00481 - 0.00943
Any cancers (19)	Men	40 to 44	0.00093	0.00072 - 0.00134
	Men	45 to 49	0.00142	0.00117 - 0.00204
	Men	50 to 54	0.00259	0.00214 - 0.00301
	Men	55 to 59	0.00428	0.00350 - 0.00532
	Men	60 to 64	0.00647	0.00541 - 0.00777
	Men	65 to 69	0.00926	0.00773 - 0.01108
	Men	70 to 75	0.01251	0.01033 - 0.01519
	Women	40 to 44	0.00156	0.00128 - 0.00203
	Women	45 to 49	0.00205	0.00174 - 0.00255
	Women	50 to 54	0.00261	0.00221 - 0.00307
	Women	55 to 59	0.00359	0.00299 - 0.00436
	Women	60 to 64	0.00473	0.00396 - 0.00568
	Women	65 to 69	0.00580	0.00495 - 0.00684
	Women	70 to 75	0.00688	0.00579 - 0.00828
Hemorrhagic stroke (19)	Men	40 to 44	0.00038	0.00013 - 0.00063
	Men	45 to 49	0.00060	0.00016 - 0.00105
	Men	50 to 54	0.00086	0.00037 - 0.00135
	Men	55 to 59	0.00115	0.00033 - 0.00197
	Men	60 to 64	0.00148	0.00062 - 0.00235
	Men	65 to 69	0.00186	0.00042 - 0.00330
	Men	70 to 75	0.00231	0.00086 - 0.00376
	Women	40 to 44	0.00046	0.00014 - 0.00078
	Women	45 to 49	0.00071	0.00017 - 0.00126
	Women	50 to 54	0.00095	0.00035 - 0.00155
	Women	55 to 59	0.00117	0.00031 - 0.00204
	Women	60 to 64	0.00155	0.00063 - 0.00248
	Women	65 to 69	0.00209	0.00052 - 0.00367
	Women	70 to 75	0.00258	0.00094 - 0.00423
Non-CVD mortality (19)	Men	40 to 44	0.00154	0.00139 - 0.00169
	Men	45 to 49	0.00194	0.00170 - 0.00217
	Men	50 to 54	0.00287	0.00252 - 0.00322
	Men	55 to 59	0.00441	0.00390 - 0.00493
	Men	60 to 64	0.00662	0.00584 - 0.00741
	Men	65 to 69	0.00984	0.00867 - 0.01101
	Men	70 to 75	0.01496	0.01304 - 0.01688
	Women	40 to 44	0.00080	0.00073 - 0.00088
	Women	45 to 49	0.00115	0.00103 - 0.00126
	Women	50 to 54	0.00172	0.00154 - 0.00189
	Women	55 to 59	0.00273	0.00244 - 0.00303
	Women	60 to 64	0.00447	0.00398 - 0.00498
	Women	65 to 69	0.00693	0.00613 - 0.00774
	Women	70 to 75	0.01106	0.00962 - 0.01249
Renal dysfunction (21)	Men	All ages	0.00074
	Women	All ages	0.0006
Hepatic dysfunction (21)	Men	All ages	0.00244
	Women	All ages	0.00246
Cataracts(21)	Men	All ages	0.0099
	Women	All ages	0.01764
Myopathy (21)	Men	All ages	0.00036
	Women	All ages	0.00032
Headache/nausea (22)	Both sex	All ages	0.04085
Treatment discontinuation (22)	Both sex	All ages	0.03518

• Weights of outcomes preferences

To determine the relative importance and preferences for different statin-related outcomes, we have previously conducted a study using the Best–Worst Scale (BWS) method in a population covered by health centres. The study involved 1085 participants, including 913 members of the general population (486 women) and 172 healthcare providers from both urban and rural primary healthcare centres in Iran [19]. Those from the general population were 40–75 years old without a history of CVD events. We used different methods to derive the quantitative weight of each outcome; the results of the conditional logistic model (log-odds) were used in the main analysis of this study, and the results of the best–worst scores were used in the sensitivity analysis (Table 2). We found that preferences were similar between rural and urban areas and among healthcare providers and the general population, with overlapping uncertainty intervals [19]. In the current study, we used the results obtained for the general population to determine the severity of the consequences when assessing the benefit-harm balance index of statins.

Table 2.

Preference weights for statin-related outcomes from an Iranian survey (18)

Outcome	Log-Odds	Best–Worst Score
Myopathy	0	0
Renal dysfunction	3.27	0.43
Hemorrhagic stroke	4.52	0.63
Hepatic dysfunction	2.93	0.43
Type 2 diabetes	3.18	0.44
Any cancer	5.76	0.89
Cataract	2.93	0.43
Headache/nausea	0.36	0.05
Treatment discontinuation	0.80	0.12
CVD	3.69	0.52

Statistical analysis

To estimate the expected number of events without statin use per 1000 persons over 10 years, we first considered the baseline risk for each outcome and the competing risk for non-CVD death using Eq. 1. Then, using the same model, but taking into account the estimates of the effect of statins, we calculated the number of expected events for each outcome in a period of 10 years per 1000 statin users using Eq. 2.

We then calculated the attributable absolute event for each outcome using Eq. 3, which represents the difference in expected events with and without statin use. To yield a single benefit–harm index, we weighted the differences by their respective preference weights estimated in the previous part of “Weights of outcomes Preferences” and summed them using Eq. 4.

$$\begin{array}{c}{N}_i,\mathrm{no}\mathrm{statin}=1000\times [\frac{{\mathrm{I}}_i}{({\mathrm{I}}_i+M)}]\times [1-{\mathrm{e}}^{-10\times ({\mathrm{I}}_i+M)}]\hfill \\ N_i,\mathrm{no}\mathrm{statins}:\mathrm{Expected}\mathrm{number}\mathrm{of}\mathrm{events}\mathrm{without}\mathrm{statin}\mathrm{use}\hfill \\ I_i:\mathrm{Baseline}\mathrm{risk}\mathrm{for}\mathrm{each}\mathrm{outcome}\hfill \\ M:\mathrm{Competing}\mathrm{risk}\mathrm{for}\mathrm{non}-\mathrm{CVD}\mathrm{death}\hfill \end{array}$$ 1

$$\begin{array}{c}\mathrm{Ni},\mathrm{statin}=1000\times \left[\frac{R{R}_{\mathit{i}}\times I_{\mathit{i}}}{(R{R}_{\mathit{i}}\times I_{\mathit{i}}+M)}\right]\times \left[1,,-,,e^{-10\times (R{R}_{\mathit{i}}\times I_{\mathit{i}}+M)}\right]\hfill \\ N_i,\mathrm{statins}:\mathrm{Expected}\mathrm{number}\mathrm{of}\mathrm{events}\mathrm{with}\mathrm{statin}\mathrm{use}\hfill \\ I_i:\mathrm{Baseline}\mathrm{risk}\mathrm{for}\mathrm{each}\mathrm{outcome}\hfill \\ M:\mathrm{Competing}\mathrm{risk}\mathrm{for}\mathrm{non}-\mathrm{CVD}\mathrm{death}\hfill \\ R{R}_i:\mathrm{Treatment}\mathrm{effect}\mathrm{of}\mathrm{statins}\mathrm{on}\mathrm{each}\mathrm{outcome}\hfill \end{array}$$ 2

$$\mathrm{Attributable}\mathrm{absolute}\mathrm{event}=N_{\mathit{i}},\mathrm{no}\mathrm{statins}-N_{\mathit{i}},\mathrm{statins}$$ 3

$$\begin{array}{c}\mathrm{Benefit}-\mathrm{harm}\mathrm{index}=\sum [(N_{\mathit{i}},\mathrm{no}\mathrm{statins}-N_{\mathit{i}},\mathrm{statins})\times W_{\mathit{i}}]\hfill \\ W_{\mathit{i}}:\mathrm{preference}\mathrm{weight}\hfill \end{array}$$ 4

A negative result indicates that harms outweigh benefits, while a positive result indicates that benefits outweigh harms.

We accounted for uncertainty in the input parameters by repeating the analysis 100,000 times for each subgroup (age- and sex-specific groups) with resampling of the parameters from normal distributions using the mean estimates of baseline risks, log risk ratio, and preference weights and their standard errors.

We measured the probability of net benefits as the proportion of repetitions with a positive benefit–harm index, which may take any value between 0 and 100%. We considered statins to have net benefits if the probability of the positive benefit–harm index was at least 60% and net harms if this probability was lower than 40% (assuming an arbitrary 10% probability of uncertainty instead of defining the cutoff at exactly 50%, as done previously) [5].

The probabilities were computed across the CVD risks 1%-40% as a plausible range considered as the treatment threshold [24]. All analyses and simulations were performed using R statistical software version 3.6.1 and STATA version 14.

Results

The baseline risks (annual rate) of the statin-related outcomes for the Iranian population are presented in Table 1. For instance, the incidence rate of type 2 diabetes ranged from 5 to 10 per 1000 subjects annually in different age and sex categories. Statins increase these rates by 9% (Supplementary Table 1). Furthermore, the annual rate of non-CVD mortality, as a competing risk, ranged around 1 to 10 per 1000 subjects. Considering these statistics and based on Eqs. 1 and 2, the number of incident diabetes per 1000 subjects during 10 years is estimated at 50–99 cases without and 54–108 cases with statins in different age and sex categories (Supplementary Tables 2 and 3). The number of CVD events with and without statins was also estimated in different CVD risks from 1 to 40% (Supplementary Tables 2 and 3). Finally, employing the weights provided in Table 2 and considering all outcomes jointly (using Eqs. 3 and 4), we calculated the benefit-harm balance for different age and sex categories. After accounting for uncertainty in the input parameters and repeating the analysis 100,000 times, Table 3 shows the mean of age- and sex-specific benefit-harm balance indices (the results are shown just for CVD risk from 1 to 25%). After the risk thresholds of 17–21% in men and 23–25% in women, the mean of benefit-harm balance indices resulted in positive results. Figure 1 however illustrates the probability of achieving a positive net benefit among 100,000 repetitions by age, sex, and level of CVD risk. These analyses show that the probability of a positive net benefit surpasses 60% at the risk thresholds of 18–23% and 24–28% in men and women, respectively, suggesting that these risk thresholds may be appropriate for initiating statin therapy for primary prevention of CVD.

Table 3.

Mean benefit-harm balance indices for statin therapy for primary prevention of cardiovascular diseases per 1000 people by age, sex, and 10-year cardiovascular risk categories

Men
Age, y	70–75	-157	-149	-140	-131	-123	-114	-105	-97	-89	-80	-72	-64	-56	-48	-40	-32	-24	-17	-9	-2	6	13	21	28	35
	65–69	-157	-148	-139	-130	-121	-112	-103	-95	-86	-78	-69	-61	-53	-45	-36	-28	-20	-13	-5	3	10	18	25	33	40
	60–64	-154	-145	-136	-127	-118	-109	-100	-91	-83	-74	-66	-57	-49	-40	-32	-24	-16	-8	0	8	15	23	30	38	45
	55–59	-151	-142	-132	-123	-114	-105	-96	-87	-78	-70	-61	-53	-44	-36	-27	-19	-11	-3	5	13	21	28	36	43	51
	50–54	-146	-136	-127	-118	-109	-100	-91	-82	-73	-64	-55	-47	-38	-30	-21	-13	-5	3	11	19	27	35	42	50	57
	45–49	-141	-131	-122	-113	-103	-94	-85	-76	-67	-59	-50	-41	-33	-24	-16	-7	1	9	17	25	33	41	48	56	63
	40–44	-136	-126	-117	-107	-98	-89	-80	-71	-62	-53	-45	-36	-27	-19	-10	-2	6	14	22	30	38	46	54	61	69
	CVD risk (%)	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25
Women
Age, y	70–75	-195	-186	-177	-168	-159	-151	-142	-133	-125	-116	-108	-100	-91	-83	-75	-67	-59	-52	-44	-36	-29	-21	-14	-6	1
	65–69	-197	-188	-179	-170	-161	-152	-143	-134	-125	-117	-108	-100	-91	-83	-75	-67	-59	-51	-43	-35	-27	-20	-12	-5	2
	60–64	-198	-189	-479	-170	-161	-152	-143	-134	-126	-117	-108	-100	-91	-83	-75	-66	-58	-50	-42	-34	-27	19-	-11	-4	4
	55–59	-197	-188	-179	-169	-160	-151	-142	-133	-124	-116	-107	-98	-90	-81	-73	-65	-56	-48	-40	-32	-24	-17	-9	-2	6
	50–54	-194	-184	-175	-166	-156	-147	-138	-129	-120	-112	-103	-94	-86	-77	-69	-60	-52	-44	-36	-28	-20	-12	-5	3	10
	45–49	-189	-179	-170	-161	-151	-142	-133	-124	-115	-106	-98	-89	-80	-72	-63	-55	-47	-39	-31	-23	-15	-7	1	8	16
	40–44	-182	-173	-163	-154	-145	-135	-126	-117	-108	-100	-91	-82	-74	-65	-57	-48	-40	-32	-24	-16	-8	0	8	15	23
	CVD risk (%)	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25

Benefit-harm balance index was calculated based on Eq. 4 in the text. Each cell shows mean from the 100,000 repetitions considering uncertainty in the input parameters. Negative results indicate that harms outweigh benefits, zero indicates harms equal benefits and positive results indicate that benefits outweigh harms

Fig. 1.

Probability (%) of net benefit of statin therapy for primary prevention of cardiovascular diseases by different age, sex, and 10-year cardiovascular risk categories. The probability of net benefit is the proportion of repetitions with a positive benefit–harm index. Each cell shows percent from the 100,000 repetitions considering uncertainty in the input parameters. We considered statins to have net benefits if the probability of the positive benefit–harm index was at least 60% (Green cells) and net harms if this probability was lower than 40% (Red cells). In probabilities 40% to 60%, harms are considered to be equal benefits (Yellow cells)

In the sensitivity analysis, we employed best–worst scaling scores to weight outcomes. Our findings suggest that the risk thresholds for initiating statins could be 19–24% in men and 25–29% in women (Supplementary Fig. 1).

Discussion

In this study, the research question was the appropriate risk threshold for the use of a low and moderate dose of statins for the primary prevention of CVD in people aged 40–75 years who had no history of CVD. The benefit-harm balance index was calculated for age, sex and different risk groups for the occurrence of CVD over 10 years. This index was used to determine the 10-year risk threshold for the administration of statins so that in values above that threshold, the probability of statin benefit is greater than its harm. Based on the results, the risk threshold for prescribing statins for the primary prevention of all CVD events, a range of angina to CVD mortality, was between 18–23% in men and 24–28% in women.

Most cardiovascular disease prevention clinical guidelines recommend the use of statins to prevent Hard-CVD (including myocardial infarction, fatal coronary heart disease and fatal or non-fatal stroke) at the risk thresholds of 7.5–10% [7, 8, 25, 26]. In the present study, it was shown that the net benefit of statins for the prevention of all CVDs including Hard-CVD and also angina and heart failure ranged from 18 to 28% based on different age-sex groups. Based on a large population-based cohort study in Tehran, the 10-year incidence of Hard-CVD, including only myocardial infarction, fatal CHD and fatal or non-fatal stroke, is 9.4% and the 10-year incidence of Total-CVD, including unstable angina and revascularization plus hard outcomes, is 21.0% in men; the corresponding values for women are 4.1% and 12%, respectively [27]. As a result, in our community, the risk of Hard-CVD is about half of the risk of Total-CVD in men and one-third of that in women. On the other hand, based on our general population’s preferences, the severity of myocardial infarction and severe CVDs is much higher than the severity of unstable angina; for example, moderate myocardial infarction ranked five versus the rank of ten for unstable angina, among 13 different outcomes of statin therapy [19]. So, the preferred risk of Total-CVD for statin therapy can be translated to half or one-third of that for Hard-CVD events. With this approach, the preferred risk for statin therapy to prevent Hard-CVD outcomes would be 9–12% and 8–10% in our men and women respectively. These results are the same as the risk thresholds of 7.5–10% considered for statin therapy in the guidelines.

Based on the previous guideline of ATP III for cholesterol treatment, those with a 10-year risk of hard coronary disease over 20% should have received lipid-lowering medication. This risk threshold was opted by the expert panel because 20% was the average probability of recurrent hard coronary events within 10 years in individuals with a history of hard coronary disease [28]. After that in the ACC/AHA guidelines, this risk threshold was decreased to 7.5% because statin’s net benefit was substantial [29]. Although there was controversy on this new risk threshold, we showed in our previous work that using this cutoff in our community could result in a net number need to treat below 50 [27]. However, data regarding the harm and benefit of treatment from the patient’s point of view were lacking, the point that has been emphasized in the Evidence-Based Medicine Triad by Sackett DL, et al. [30]. Thus in the current study, we tried to address this issue and found that generally, patients prefer to be treated with risks above 20–30% for Total-CVD which is translated to risks above 10% for Hard-CVD. It should be noted that this threshold is an average for the general population and in the clinic and personalized medicine, it could be different for different individuals based on risk communication between physician and patient, the point that has been emphasized in the new guidelines, especially for adults at borderline (5% to < 7.5%) and intermediate (≥ 7.5% to < 20%) risk to consider additional risk-enhancing clinical factors in clinician–patient risk discussion [31].

Clinical guidelines emphasize the benefits of statins and it seems that the potential harms of statins have little effect on the recommendations [32]. In the present clinical guidelines, statins are recommended for the prevention of CVD at a risk threshold of 7.5 to 10%. The problem with such an approach is that the risk of CVD increases with age, and the justification for taking statins increases with age because statins can prevent more events in the elderly who are at risk for CVD [25]. However, the probability of adverse outcomes also increases with age, which makes it more challenging to calculate the benefit-harm balance of statins. The results of the present study showed that the 10-year risk thresholds for CVD, in which the benefits of administering statins are greater than their potential harms, are higher in the elderly than in the young. The same results came from the studies of Yebyo HG. et al. [5, 6] Some older people, even in the absence of other risk factors for CVD, are eligible for statins. On the other hand, young people may not have a high calculated risk of disease, even with relatively high-risk factors [24]. In other words, an increase in age relates to increasing risk thresholds the point which has been considered in the 2021 European Society of Cardiology (ESC) Guideline on cardiovascular disease prevention in clinical practice. In this guideline, healthy people were categorized as very high CVD risk when risk based on the model of SCORE2 > 7.5% for age under 50, > 10% for age 50–69, and > 15% for age > 70 [14].

We did the same method that Yebyo et al. used to determine the global variation of risk thresholds for initiating statins for primary prevention of CVD, however, we used our own weights of outcomes preferences besides local incidences for these outcomes. Yebyo, et al. [6] showed that the risk threshold for prescribing statins for primary prevention of CVD among 186 countries depending on age groups was 14–22% in men and 19–24% in women. In their study, the risk threshold for prescribing statins for Iran was between 14–17% in men and 19–21% in women, which is lower than the thresholds obtained in the present study (18–23% in men and 24–29% in women.). They elicited outcome preferences from potential statin users using the best–worst scaling survey in Switzerland and Ethiopia and applied the results to all 186 countries [5, 6, 23]. They noted that the preferences were very similar between Switzerland and Ethiopia and thus they took the average estimates to apply across countries.

More importantly, since the outcome preferences elicited by Yebyo et al. include all cardiovascular events from mild outcomes (e.g. unstable angina) to severe ones (severe myocardial infarction and stroke and fatal events), the obtained risk thresholds are due to Total-CVD outcomes, not Hard-CVDs. Most clinical prediction models for CVDs calculate the risk of Hard-CVD events (e.g. Globorisk, ACC/AHA pooled cohort equation, the revised WHO CVD risk charts and SCORE2) and most guidelines consider Hard-CVD events as their main purpose for prevention [12, 14]; except NICE guideline for CVD risk assessment uses the QRISK2 prediction model which considers all CVD events including angina [33]. So Yebyo et al.’s conclusion that statins provide net benefits at higher risk thresholds than are provided by most current guidelines may not be justified because the incidence of Total-CVD is around 1.5 to three times of Hard-CVD [27, 34].

The method used in our study has never been applied to the standards of existing international lipid guidelines, however, we were able to look at the matter with a different approach which has been firstly used for cancer medicines and then applied for CVD prevention; [5, 17] of note, the primary prevention with statins took decades, once the cancer prognosis is poor and always leads to death for a short period of time. Despite these differences and limitations, we reached the same risk thresholds for statin therapy as recommended by current guidelines. Moreover, this study did not consider the costs associated with statin use, however, in our previous work, we demonstrated that preventive strategies, including statin therapy, would be cost-effective for individuals with a 10-year hard CVD risk of 10% or higher, which aligns with the results of the current study [35].

The present study has limitations that should be considered in the interpretation of the results. The weight of people’s preferences for different age and sex groups was considered the same. In our previous work, we did not see a statistically significant relationship between the expressed preferences with age, sex and place of residence [19], a point which is in line with other studies in the field [23, 36]. All diabetic subjects and those who have LDL =  > 190 mg/dl should take statin regardless of their CVD risk [31]. Since the information required for calculating the benefit-harm index is not available for the general population excluding diabetic subjects and those who have LDL =  > 190 mg/dl, we may have met a kind of underestimation for the risk threshold. On the other hand, statins reduce all-cause mortality as well [37] and this point increases their benefits and this point decreases the optimum risk thresholds. It is important to note that, the current analysis focuses on determining the Benefit-Harm index of statin therapy solely based on different levels of CVD risk as an indication for statin therapy; the harm-benefit profile of other indications for statin therapy is beyond the scope of this study. However, the baseline risk of both CVD and non-CVD outcomes, influenced by any reason such as the prevalence of diabetes, is a key consideration that may vary across different populations and impact the Benefit-Harm index. Nonetheless, because diabetes and cholesterol levels are considered in risk scoring, a subset of high-risk individuals based on risk scoring are those with diabetes or high cholesterol. Thus, most high-risk individuals in the general population who are eligible for statin therapy are selected using risk scoring.

We utilized information on main serious outcomes including type 2 diabetes, cancers, hemorrhagic stroke, and total non-CVD death from the Global Burden of Disease (GBD) study, which is the primary reliable source of health outcomes worldwide. The other outcomes were not as serious and critical as the main outcomes and were not available in the GBD study. Finally, the current study considered a low to moderate dose of statins, which is typically prescribed for primary prevention of CVD and used in most clinical trials. However, in practice, we may observe underuse of statin therapy [38]. This underscores a general limitation in applying the results of clinical trials directly to real-world practice, as there can be a difference between the efficacy observed in trials and the effectiveness achieved in actual clinical settings.

Conclusion

Since clinical guidelines are primarily based on studies conducted in developed countries, our research validated these guidelines within the context of our population. The reduction in total CVD events (encompassing a spectrum of outcomes from angina to CVD mortality) from statin therapy outweighs the adverse events for individuals with a 10-year total CVD risk of 18–23% in men and 24–28% in women across different age groups. Because the main outcome in clinical guidelines is hard CVD including only myocardial infarction, fatal coronary heart disease and fatal or non-fatal stroke, our results can be translated to the hard CVD risk thresholds of approximately 7.5% to 10% for both sexes, as recommended by most guidelines for statin therapy. Also, older age groups may require higher risk thresholds for statin therapy in primary prevention of CVD within public health. To effectively apply the findings of this study, it is essential to determine the 10-year CVD risk in individuals using valid and well-calibrated predictive models tailored to our population.

Abbreviations

CVD

Cardiovascular disease

LDL

Low-density lipoprotein

BWS

Best–worst scaling

WHO

World Health Organization

ACC

American College of Cardiology

AHA

American Heart Association

GBD

Global Burden of Disease

Authors' contributions

All authors, DK, HS, HRB, FH, EWS, MW and GD contributed to the study conception and design. Material preparation and data collection were performed by DK, HS and HRB. HS and DK carried out data analysis, and HRB, FH, EWS, MW and GD provided critical comments on that. The first draft of the manuscript was written by HS and DK and HRB, FH, EWS, MW and GD commented on the manuscript. All authors read and approved the final manuscript.

Funding

This study is a part of a project funded by the National Institute of Medical Research Development (NIMAD) (grant no.964114) in Iran, as well as a seed grant from the Lown Scholar Program in the Department of Global Health and Population at Harvard T.H. Chan School of Public Health. The funding bodies had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ethics approval for this study was obtained from the Ethics Committee of the Iran University of Medical Sciences (IR.IUMS.REC.1397.1073). Since publicly accessible data were used, informed consent were not required for this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Supplementary Information

Additional file 1. This contains supplementary tables and figure, including an effect estimates of statins, number of expected events per 1000 subjects during 10 years for each of the statin-related outcomes and probability of net benefit of statin therapy.