In total, 363 AI patients were found, and 327 of these fulfilled the inclusion criteria (149 PAI [46%], 178 SAI [54%]). We included 216 women and 111 men, with most of the patients being between 31 and 50 years (33%) and between 51 and 69 years (32%). Twenty-one percent were 70 years and older. Mean age was 53 years (range 18–94). The vast majority of patients (n = 264) were on conventional hydrocortisone, 36 patients were treated with modified released hydrocortisone (Plenadren®) and 8 patients took prednisolone. However, 19 participants were substituted with a fixed combination of conventional hydrocortisone, modified released hydrocortisone resp. prednisolone. Interestingly, dexamethasone was given additionally to Plenadren® or prednisolone in two of these cases.

Mean dosage of GC (dosing equivalents relative to hydrocortisone) was 20.1 mg with a standard deviation of 9.9 mg. Two-hundred-thirty-six patients took less than 30 mg (79%), whereas 48 AI patients took 30 to 35 mg (15%) and 21 participants took even more than 35 mg (6%). Ninety-two PAI patients were taking fludrocortisone (62%) with a mean dose of 0,09 mg.

Of the 327 included patients, 299 were found to have one or more CVR factors. However, in 28 patients the risk factors were not sufficiently assessed, meaning they did not have any data for one or more risk factors. We found the following CVR factors: arterial hypertension, arteriosclerosis, coronary heart disease, peripheral artery disease, stroke, diabetes mellitus, hyperlipoproteinemia, hyperuricemia, obesity, sleep apnea syndrome, coagulation disorder, aneurysm, smoking, positive family history for the conditions mentioned above, as well as elevated values of lipids, uric acid, HbA1c, BMI or blood pressure. In total we identified 27 risk factors.

Only one patient did not have any risk factors. Most AI patients (n = 299) were found to have at least one risk factor (91%).

Between the three participating locations, there were significant differences regarding the evaluation of CVR factors.

On average, the patients in Hamburg were found to have about ten missing risk factors, whereas in Bremen about two factors were missing, and in Leer approximately three. There was a significant difference between Hamburg and Leer (p < 0.001) and Hamburg and Bremen (p < 0.001).

Elevated blood pressure (BP) was the leading global contributor to premature death in 2015 and is above all related to cardiovascular events such as stroke, myocardial infarction, heart failure, peripheral artery disease, and end-stage renal disease [16, 17]. Hypertension is defined as office systolic BP values > 140 mmHg and/or diastolic BP values > 90 mmHg [18].

In 91 patients (27.8%) arterial hypertension was documented with 44% (n = 40) of them still having a higher BP in one measurement. Almost half of the AI patients with known hypertension showed normal or low BP (n = 46) and five patients had no documented BP (5.5%) at all (Fig. 1).

Arterial hypertension and blood pressure. a Patients with known arterial hypertension, b patients without known arterial hypertension, c patients without information about arterial hypertension

In patients, classified as not having arterial hypertension (n = 236), interestingly 21.6% showed elevated BP (n = 51). For 168 patients normal or low BP was documented (71.2%). Similar to patients with known arterial hypertension, a relevant proportion of this patient group (7.2%) did not provide any BP value (n = 17).

Concerning medication, three-quarters of the AI patients with known hypertension did not take any antihypertensive drugs according to the medical records. Furthermore, in patients who currently demonstrated high BP measurements, there were 65.9% who did not take any antihypertensive medication. Antihypertensive dose adjusting was performed by the treating family physicians in almost all cases.

Overweight and obesity are defined as abnormal or excessive fat accumulation that may impair health. Obesity is a body mass index (BMI) greater than or equal to 30 kg/m2 [19]. A BMI between 25 kg/m2 and 29,9 kg/m2 is considered overweight. Typical complications comprise type 2 diabetes, hypertension, fatty liver disease, and obstructive sleep apnea [20]. The distribution of BMI and obesity is shown in Fig. 2.

Overview of BMI distribution in AI patients. a distribution of BMI, b distribution of obesity classes in AI patients with obesity

In total, 163 patients were overweight or obese. The threshold for obesity was reached by 78 of the AI patients with 13 having a BMI of 40 or higher.

About 80% of all AI patients with a documented diagnosis of obesity in the medical records had an elevated BMI (>25 kg/m2, n = 91), but only 60.3% were in fact obese (BMI > 30 kg/m2). However, 17 of these cases (15.3%) had no documented BMI at all. Three of the patients with documented diagnosis of obesity had BMI values in the normal range (almost 3%) (Fig. 3).

Obesity and BMI. a with diagnosis of obesity, b patients without diagnosis of obesity

Surprisingly, 5% of all AI patients who were not classified as suffering from obesity, were in fact obese (n = 11). Only 36.1% of this patient group had BMI values in the normal range (n = 78). The threshold for overweight was reached by 72 patients (33.3%). In 66 cases weight measurements were missing (30.6%). No weight measurements were documented for 66 patients (30.6%).

Patients with elevated BMI were often found to have more than one CVR factor. Merely five of these patients had elevated BMI as the only risk factor. Fifty-three of these patients were also known to have arterial hypertension (33%). Ninety-one patients with elevated BMI showed increased lipid levels (56%) as well.

Patients with obesity did not get any medical therapy to reduce weight except lifestyle modifications.

The 48 mmol/mol (6.5%) HbA1c threshold was used to monitor blood glucose. Target HbA1c levels for patients with type 1 diabetes mellitus (T1DM) are between 6.5 and 7.5% [21]. The target levels for type 2 diabetes mellitus (T2DM) were recently defined individually between 6.5 and 8.5% [22]. For this study, we used the 7.5% threshold for both types. Typical complications of diabetes mellitus (DM) include cardiovascular, renal, peripheral vascular, ophthalmic, hepatic, or neurological diseases [23].

Although 46 patients suffered from DM, more than 1/4 (n = 12) still had increased HbA1c values (Fig. 4).

Diabetes mellitus and HbA1c levels. a patients with known diabetes mellitus, b patients without known diabetes mellitus

No DM was documented for 281 patients. However, in 17.5% of these cases, an elevated HbA1c was found (n = 49). The majority had prediabetes, only in one case the threshold for DM was reached.

Only four of the diabetic patients took antidiabetics. There were eleven patients with known DM still having HbA1c over 7.5% who did not take any antidiabetic medication. Interestingly, there were also four patients without known DM and no data for HbA1c who did take antidiabetic drugs.

Hyperlipidemia is known to be an important CVR factor [24]. The local laboratory threshold for total cholesterol was 200 mg/dl. The most important subgroup is LDL cholesterol as it is the major risk factor in the formation of atherosclerotic plaques [24]. LDL cholesterol is clearly elevated if it is 155 mg/dl or higher. Triglycerides act as a predictive marker for cardiovascular events [25]. They are marked as elevated when over 150 mg/dl. The role of Lp(a) has become more important over the last few years; it is considered to be an independent risk factor for CVD [26, 27]. If Lp(a) is 72 nmol/l or higher it is regarded as elevated.

Concerning hyperlipidemia, about 55% of all AI patients had at least one elevated lipid level (n = 181). Only three patients demonstrated lipids in the normal range in each subgroup (total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, and Lp(a)). Hundred-fifty of all AI patients were not adequately assessed, meaning that at least one value was unavailable. However, the existing lipid levels were normal. Out of the 181 patients with at least one elevated lipid, only 57 had documented hyperlipidemia. Seventeen patients with known hyperlipidemia had not been screened for elevated lipids (22.4%) (Fig. 5).

Hyperlipidemia and lipid levels. a patients with known hyperlipidemia, b patients without known hyperlipidemia

Only 17% of AI patients had data for Lipoprotein a (Lp(a)). Surprisingly no AI patient in Leer and only nine AI patients in Bremen had data for Lp(a).

In total, 126 patients had elevated total cholesterol levels (38.5%). There were six patients with data for total cholesterol but no further evaluation of the subgroups. In addition to elevated cholesterol levels, 53 AI patients also exhibited high values for LDL cholesterol.

Surprisingly, only 18 patients with known hyperlipidemia took lipid-lowering agents (23.7%). Moreover, in patients with at least one elevated lipid level, only 17.1% were taking lipid-lowering agents (n = 31).

Hyperuricemia is closely related to CVD with higher levels of uric acid (UA) being a risk for CVD [28]. It is defined as an elevation of serum UA (>6 mg/dL in women and > 7 mg/dL in men) [29]. We used the 5.7 mg/dL threshold as this was the laboratory reference. The underlying mechanisms of the increase in CVR are still not completely understood and are discussed controversially [30].

Only 21 of the AI patients were known to suffer from hyperuricemia (6%). Of these patients, five did not have data for uric acid (24%) (Fig. 6).

Hyperuricemia and uric acid levels. a patients with known hyperuricemia, b patients without known hyperuricemia

In 306 AI patients, no hyperuricemia was known. Though, 37 of these patients demonstrated elevated UA (12.1%). A big proportion (53.6%) did not have any measurements for UA (n = 164).

Unexpectedly, no patient with known hyperuricemia took uricostatic drugs but seven of the patients who, according to the documentation, did not know about their hyperuricemia.