Time trends (2001–2019) and sex differences in incidence and in-hospital mortality after lower extremity amputations among patients with type 1 diabetes in Spain

Lower extremity amputation (LEA) is a major complication of diabetes and is associated with low quality of life and higher risk of short-term mortality [1]. People with type 1 diabetes mellitus (T1DM) have a higher risk of LEA [2]. A recent study showed a 40-fold greater risk of amputation in patients with T1DM than in the general population [3]. The risk of LEA in patients with T1DM is a result of the combination of several conditions whose key contributing factors include ageing, male sex, cardiovascular comorbidities, hyperglycaemia, hypertension, and hyperlipidaemia [2, 4].

Rates of LEA among adults with diabetes are an important index of comprehensive diabetes care. Population-based studies in Spain showed that the number of T1DM-related LEAs decreased between 2001 and 2008 and then remained stable through 2012 [5, 6]. A Swedish study also reported a decrease in the frequency of amputations in patients with T1DM between 2017 and 2019 [4]. Similar findings have been reported in Denmark [7]. However, a study performed in the United States found an increase in diabetes-related LEAs, predominantly among younger patients (aged 18–44 years) and middle-aged patients (45–64 years) [8].

Sex differences may play an active role in incidence and outcomes among patients with diabetes following LEA. Several studies indicate that male sex is associated with a higher incidence rate than female sex [5, 6, 9], although female sex is an independent predictor of surgical site infection and in-hospital mortality (IHM) following LEA [6, 10]. Data regarding the results of hospitalization following LEA among men and women with T1DM are scarce.

Reliable epidemiological data on trends related to LEA, sex differences, and associated mortality are useful for assessing whether the quality of care for people with T1DM has improved and for informing health policy makers and healthcare providers of where additional resources may be required to fill the care gaps. Therefore, we used administrative data from an entire country over an 18-year period to examine trends in the incidence, clinical characteristics, and in-hospital outcomes of LEA procedures (major and minor) among patients hospitalized with T1DM. We also attempted to identify possible sex differences. Furthermore, we investigated the variables associated with IHM according to the type of LEA.

A total of 6,011 non-traumatic amputations (3,992 minor and 2,019 major, ratio 1.97) were identified in patients aged ≥ 18 years with T1DM in Spain between 2001 to 2019. The proportion of women was 27.31% (1,642).

This nationwide registry and population-based observational study showed a decline in hospital admissions for minor and major LEAs in patients with T1DM from 2001 to 2009–10, with no subsequent changes. However, from 2017 to 2019, the incidence of major LEA seems to have increased among men with T1DM. The incidence rates of minor and major LEA in men with T1DM were higher than in women with T1DM. Female sex is a predictor of IHM in patients with T1DM following major LEA. In the fully adjusted model, women with T1DM had a 37% higher adjusted risk of dying following major LEA than men with T1DM.

According to our database, trends in incidence decreased significantly between 2001 and 2009 for minor LEA (9.55% per year) and between 2001 and 2010 for major LEA (13.39% per year) and subsequently remained stable for major LEA. However, for minor LEA the incidence increased, albeit not significantly, between 2010 and 2019. These findings have been reported elsewhere in Europe [4, 7]. In a population-based cohort study in Sweden, Hallstrom et al. [4] found that the frequency of amputation in patients with T1DM decreased between 1998 and 2019 and concluded that these changes are possibly related to an increased focus on risk factor management, advanced treatments to optimise glycaemic control, and further enhancements in the armamentarium of multidisciplinary diabetes foot clinics, including an increased use of invasive arterial reperfusion.

Although decreasing rates of LEA may reflect improved care, they may also reflect delayed presentation and early death in patients with foot ulceration or ischemia [19]. Consistent with these findings, the cross-sectional study of patients with T1DM and T2DM by Geiss et al. [8] performed in the USA found an increase in diabetes-related LEA rates between 2009 and 2015. This increment was driven by a 62% increase in the rate of minor amputations (from 2.03 [95% CI 1.83–2.22] to 3.29 [95% CI 3.01–3.57], p < 0.001) and a smaller, but also statistically significant, 29% increase in major LEAs (from 1.04 [95% CI 0.94–1.13] to 1.34 [95% CI 1.22–1.45]). Furthermore, these increases were more pronounced among the young and middle-aged groups (age 18–64 years), raising concerns that preventive efforts were not reaching certain subgroups of people with diabetes who may have different disease characteristics and needs [8]. It has been suggested that young people with diabetes have less motivation for diabetes self-care than for other life priorities and are also more vulnerable to diabetes-related distress and other mental health problems, which adversely affect disease control and increase the risk of complications [20]. Regarding this point, the non-significant increase observed in the incidence of LEA in men with T1DM in recent years should be closely monitored to identify a possible change in trend.

As we expected, our results show a high ratio of minor LEA to major LEA among patients with T1DM, possibly because of more conservative limb salvage procedures and increasingly more aggressive treatment of peripheral arterial disease in this population [21].

The results of the present study are in line with those found in the literature, namely, an increasing male predominance in minor and major LEAs [2, 4, 6, 21]. Similar sex differences were reported in a study conducted using data from the IBM Market-Scan research database among patients with T1DM and T2DM [22]. Male sex appears to be a greater risk factor for LEA than age, probably because of the differential sex distribution of risk factors, such as smoking, which were not measured. In addition, the protective role of oestrogen could lead to differences in immune system function between males and females. Finally, the biological factors of diabetic foot ulcer, peripheral vascular disease, coronary artery disease, and peripheral neuropathy might account for the significant sex difference in the amputation rates [23, 24].

While the results of the present study indicate that hypertension decreased over time, the frequency of chronic kidney disease and lipid metabolism disorders increased in patients who underwent minor and major LEA. The association between LEA and person-level risk factors in individuals with T1DM has been studied elsewhere [2, 4, 22, 25]. Hallstrom et al. [4] recently described cardiovascular comorbidities, renal dysfunction, increased HbA1c, hypertension, and smoking to be associated with an increased risk of amputation in patients with T1DM.

IHM rates for minor and major LEAs did not change significantly between 2001 and 2019. The current results reinforce those previously found in Spain for the period between 2001 and 2012 [6]. This lack of improvement over time should be investigated to identify possible explanations and implement protective measures.

During admission for major LEA, women with T1DM had a higher IHM, which remained unchanged after the multivariable regression analysis. Results on the influence of sex on IHM following LEA are contradictory. Various studies found that amputation was more frequent in men than in women [25]; however, female sex was a predictor of in-hospital mortality after LEA in a previous study in Spain [6]. Gurney et al. [26] found little difference between the sexes in terms of the adjusted risk of postoperative mortality (90 days) and concluded that female patients who undergo amputation are similar to their male counterparts in terms of the underlying risk factors for postoperative mortality.

Gender differences in glucose control in T1DM subjects, with women showing a worst glucose control possibly due to pathophysiologic differences rather than the care provided, could also contribute to the differences observed [27].

It has been repeatedly demonstrated that the presence of diabetes increases the risk burden for cardiovascular complications among females. The mechanisms involved in the excess risk for vascular complications due to diabetes in females are the object of discussion and investigation and include hormonal factors, genetic predisposition linked to sex dimorphism, and sex-related inequalities in treatment or socio-economic status, which individually or together seem to contribute to the higher incidence in females [28, 29]. Indeed, in Spain, a sex gap has been reported in the impact of diabetes on IHM after major cardiovascular events [30]. The multifactorial causes for excess risk among women with T1DM require further investigation.

In our investigation, as reported in previous studies, the comorbidities that were significant predictors for death after LEA included PAD, ischemic heart disease or chronic kidney disease [31‐34]. Morbach et al., found that the long-term survival after LEA is poor, especially among people with diabetes with concomitant PAD, renal insufficiency or the combination of both [31]. Diabetic patients with PAD, compared with those without PAD, have a higher risk of cardiovascular and all-cause mortality [32].

Interestingly, in our study neuropathy was not significantly associated with death as reported before [31]. Even more, Cascini et al. reported a protective effect of neuropathy on mortality after major LEA. We agree with these authors that a possible explanation is a differential under-reporting in hospital charts of peripheral neuropathy [33]. Indeed, the diagnosis of neuropathy is a complex process and requires expertise that is not always available [33].

The strength of our study lies in its large sample size (6,011 nontraumatic LEA procedures in patients with T1DM), coverage of an entire country (> 95% of all hospital admissions), and standardized methodology (extensively used for research in Spain, combined with the reliability of diabetes and LEA coding in the RAE-CMBD) [5, 6, 30]. Nevertheless, our study is subject to a series of limitations. Our data were obtained from an administrative database supported by the information that physicians recorded in the discharge report; therefore, the database lacks information on clinical characteristics, glycaemic control, medical treatments, and time with T1DM. Another limitation is that anonymity precludes the extraction of specific data that may affect the results (i.e., people who moved from one hospital to another could appear twice). However, Buckley et al. [35] detected high levels of agreement between hospital discharge data and medical records for LEA and diabetes and suggested that hospital discharge data are sufficiently reliable for monitoring trends in LEA in people with diabetes. In our opinion, the strengths of this study and its uniqueness clearly outweigh its limitations.

In conclusion, our data show a decrease in the incidence of minor and major LEA in men and women with T1DM and a slight but not significant increase in the incidence of major LEA in men with T1DM during the last years of the study. This possible change in the time trends must be closely followed in the future. The incidence of minor and major LEA was higher in men than in women. Female sex is a predictor of IHM in patients with T1DM following major LEA. As the study was based on administrative data, we lack relevant clinical data, thus making it necessary to take into account the effect of residual confounding. Our results should be interpreted with caution and confirmed using data from clinical studies. However, our results should be the starting point for future investigations that can confirm or reject our conclusions using more extensive and precise clinical data.