This post-hoc analysis demonstrates for the first time in a clinical trial which compared multifactorial versus standard intervention in type 2 diabetic population at high CV risk that CV prognosis and all-cause death significantly worsen with increased number of CV risk factors not reaching therapeutic goal. These results appear even more clinically interesting because they are achieved in a population undergoing primary CV prevention.

Although it is necessary to manage the CV disease burden in diabetes as effectively as possible, RCTs have not completely clarified the most effective overall therapeutic strategy. RCTs have almost always demonstrated the impact of a single CV risk factor and rarely have intensive multifactorial approaches been analysed.

A cohort study of about 270,000 type 2 diabetic patients were registered in the Swedish National Diabetes Register and matched with over 1,350,000 controls [14]. During a median follow-up of 5.7 years, subjects studied were assessed according to age categories as well as the presence of five CV risk factors (high glycated hemoglobin level, high LDL cholesterol level, albuminuria, smoking, and high blood pressure). A strict relationship between increasing number of CV risk factors not within target ranges and a higher risk of MACEs was observed. Particularly, younger diabetic subjects with multiple CV risk factors on target benefited most in terms of MACEs reduction. Notably, type 2 diabetic patients who had all five of the risk factor variables assessed within target ranges showed similar risks of death, MI, and stroke as compared with the general population. In contrast to the NID-2 study, in this very large Swedish cohort diabetic kidney disease (DKD) was either mild or absent in the majority of patients (mean estimated glomerular filtration rate (eGFR) 84 mL/min and less than 5% with macroalbuminuria). Moreover, achievement of targets for multiple risk factors was uncommon (5%), as was to be expected in an observational study. The findings from the Swedish Register are a call to action on the need for an intensive multifactorial therapeutic approach; however, according to the EUROASPIRE IV survey, a multi-drug approach to the main CV risk factors is not sufficient to achieve the goals suggested by scientific societies in the diabetic patient [15]. Hence, a multidisciplinary approach in real life and above all RCTs designed for this purpose are needed [16,17,18,19,20].

A meta-analysis of 7 RCTs did not support that intensive multifactorial intervention compared to standard of care reduced the risk of non-fatal MI, non-fatal stroke, CV disease mortality and overall mortality in microalbuminuric type 2 diabetic patients [21].

Intriguingly, the same meta-analysis showed intensive multifactorial risk factor control intervention significantly lowered blood pressure but showed a non-significant trend of reduction in HbA1c, total cholesterol, LDL cholesterol, triglyceride, and albuminuria.

Interestingly, these latter findings are partially supported by another meta-analysis of 19 RCTs in patients with type 2 diabetes. It was observed that multifactorial interventions significantly reduced the risk of non-fatal MI, but did not lower non-fatal stroke, CV disease and overall mortality [22].

Until last year, the Danish study Steno-2 and the Japanese study J-DOIT3 were the main studies which compared multifactorial intensive treatment and standard of care in type 2 diabetes [23, 24]. While the former only evaluated a microalbuminuric population, the latter analysed a diabetic population not selected for DKD. Steno-2 showed a significant reduction in MACEs and microvascular complications after an average of 7.8 years in the intensive treatment group, while the intervention group in J-DOIT3 did not reach a significant difference as compared to control group regarding fatal and non-fatal CV events after an average of 8 years. Furthermore, the post-intervention follow-up of the Steno-2 study observed a reduction in mortality in the intensive treatment arm [25].

Originally, the NID-2 trial obtained similar results as compared to Steno-2 that had an intervention phase lasting less than half (3.8 years).

But above all, these results were observed through a multicenter study on a larger population on primary prevention with patients equally distributed between the two genders, much closer to replicating the reality of type 2 diabetic patients regarding eGFR and with microvascular damage assessed by both albuminuria and DR [26]. Thus, the NID-2 study recruited a very high-risk CV population but on primary prevention, many years before recent ESC/EASD guidelines redefined the grading of CV risk in diabetes.

For decades, the guidelines of Scientific Societies have indicated HbA1c and blood glucose values as CV risk factors goals that patients with diabetes should achieve. Over the years these suggestions have often become more stringent as for the LDL cholesterol target, alongside the choice of both antihypertensive and antiplatelet drugs to use in the diabetic patients with high CV risk [3]. Assessed in their entirety, these goals are often not achieved both in the few RCTs with multifactorial treatment and in real life, both due to objective difficulty in achieving them and often because of therapeutic inertia [23, 24, 27]. Moreover, the target of single CV risk factor is often not reached in RCTs [28]. Therefore, the awareness that the achievement of a wide number of CV risk factors on target reduces CV morbidity and mortality may drive the physician and the patient to accept a multi-drug treatment, usually poorly tolerated in the daily management of diabetes. In the intensive multifactorial treatment arm of the NID-2 study the achievement of both a high percentage of subjects on target for a single risk factor and a high overall number of risk factors on target confirms that it is possible to achieve goals set out in guidelines in the clinical setting.

Our results about the association between risk factors and outcomes suggested that the risk factors might be a mediator of the intervention. Nevertheless, the association adjusted for intervention prevent the effect of other possible confounding related to itself.

This study has several strengths. Firstly, as far as we know, it is the first RCT evaluating intensive multifactorial treatment versus SoC which demonstrates that as the number of risk factors on target increases, the CV prognosis progressively improves. Secondly, these results may represent an important motivation for overcoming both the therapeutic inertia of doctors and the poor adherence of diabetics for multidrug therapy.

However, there are several limitations of the study. Firstly, post-hoc analysis results in an assessment beyond the original study design. In fact, the three analyzed groups are not homogeneous. In particular, the 4 CV risk factor group is smaller than the other two groups, this is due to the choice of variables to classify the risk factors groups.

Instead, the reason why the first two blocks were combined is to have larger groups and more robust estimates. On the other hand, this paper analyzed the same primary and secondary outcomes as in the NID-2 study, therefore the parameters analyzed are the same as in the original design of this RCT. Furthermore, despite the multicenter design of the study, the recommendation based on the study sample size was to distribute the subjects into three clusters according to the number of risk factors on target. Thus, this sub-division did not allow us to analyse the effect of achieving a single target at a time on the endpoints.

Moreover, only a minority of patients from MT group recorded data about physical activity and urinary sodium excretion as measure of dietary sodium intake. Therefore, the impact of non-pharmacological treatment on study outcomes was not analysed.

Finally, randomization by center and not by patient results in unblinded assignment, reducing both power and precision compared to an individually randomized study, and the ability to control for both known and unknown confounding variables. However, as described elsewhere, the main results of the study were adjusted for the cluster factor due to randomization. Moreover, the results were adjusted for the main variables resulted either unbalanced or clinically and statistically associated with outcome [4].