Among the 256 participants with T2DM and a new DFU in this multicenter prospective cohort study from different primary care centers in Catalonia, we found a high risk for mortality, amputations and recurrence of a new DFU. So far, similar studies to ours have been carried out in a hospital setting or in multidisciplinary foot-care centers [4,5,6,7,8,9,10,11]. These studies differ from ours, especially in the level of the healthcare system where participants were recruited, and also for the inclusion criteria, the definition of the foot ulcer, and the follow-up time, which make comparisons with our findings difficult. From the studies conducted in primary care settings, similar to ours, Boyko et al. [12] performed a study in US veterans with a follow-up period of 22 years, where all of the participants were males. The study carried out by Muller et al. [13], which assessed the annual incidence of DFU and amputations among T2DM people registered in a database of 4,500 people with different chronic conditions, used a different methodology to ours; around 677 people with diabetes per year were studied between 1993 and 1998, with a reported annual incidence of DFU and amputation of 2.1% and 0.6%, respectively, however, the authors provided very few clinical data, precluding a comparison with participants from our study.

In our cohort, we found a mortality rate of 9.7%. Three hospital-based studies have reported mortality rates in people with DFU with similar follow-up periods to our study [6,8.9]. In the study by Prompers et al. [6] with patients from 14 European hospitals, the authors reported lower mortality rates (6%). However, these people were much younger (mean age 65 years) compared with our participants (mean age 72.2 years). The other two hospital-based studies with similar follow-up periods reported mortality rates much higher than in our study. The study carried out in Germany [8] with type 1 and type 2 diabetic patients showed a mortality rate of 15.4%, while a study carried out in China by Jiang et al. [9] reported mortality rates of 14.4%. The higher prevalence of comorbidities among the people included in the study from Germany and the large number of smokers (43%) in the study from China, could partially explain these high mortality rates. As early as 1990, Apelqvist [22] warned that diabetic patients with a foot ulcer are at high risk of death. A meta-analysis performed by Saluja et al. [23] and by Brownrigg et al. [24] showed that DFU is associated with an increased risk of all-cause mortality compared to those without foot ulceration. In our study, in the adjusted model we did not find that macrovascular events (stroke, ischemic heart disease, heart failure) were associated with an increased risk of mortality in people with DFU as has been reported previously by other authors [5, 11]. In the meta-analysis conducted by Brownrigg et al. [24] the authors observed similar findings to ours regarding cardiovascular events and mortality between people with DFU and without DFU. In our study, we found an increase in mortality among women compared to men, which some authors have previously described and attributed to a greater frailty in women with DFU [25]. We also observed that ischemia and infection increased the risk of death in our multivariable analysis. These variables are well-known risk factors for a poor prognosis in subjects with DFU [26, 27]. The inverse relationship between ulcer depth and mortality may be explained by the difficulty in measuring ulcer depth on many occasions [17, 28].

Regarding amputations, we observed 31 (9.7%) events during the follow-up period, much higher than the annual incidence of this event (0.6%) reported in the cohort study by Muller et al. [13]. In similar hospital-based studies, the incidences of amputation events ranged from 0.05 to 19% [6, 11, 29]. Significant variation exists in the incidence of lower limb amputation even within the same country [30, 31]. We found that the association between amputation and retinopathy was consistent throughout all the models performed. This association was previously described in other studies [32] and indicates the importance of performing ophthalmic examinations in patients with DFU and increasing foot care at the moment of diagnosis of diabetic retinopathy.

Overall, 29.2% of participants in our study experienced a recurrence of a new foot ulcer during the follow-up. There is great variability between the studies regarding this outcome, ranging, for example, from 25% in a study by Muller et al. [13], 32% in a study by Jiang et al. [9] and 43% in a study by Winkley et al. [5]. Age was negatively associated with DFU recurrence, however it is possible that some of the older adults in our study died before the DFU recurrence, and therefore this result should be interpreted with caution. The relationship between the history of ulcers or amputation and recurrence of ulcers found in the univariate analysis disappeared in the multivariable analysis, however this is in contrast to previous studies where both variables (history of ulcers or amputation) have been reported to be poor prognostic factors for DFU recurrence [33].

Regarding DFU healing, we found that in 73.8% of participants the index ulcer healed with or without amputation. Similar results were reported previously by other authors [4, 6, 7, 34]. No relationship was observed between the ulcer's depth, extension and healing, adjusting for different variables, in contrast to what has been reported by other authors [26, 34, 35]. Our analysis highlights that infection is the main variable that interferes with healing. In a study by Prompers et al. [26] no differences were observed for major amputation or healing rate between neuropathic ulcers with and without infection, although infection was a risk factor for minor amputation. In contrast, infection was an independent predictor of poor outcome in patients with peripheral arterial disease, but the prevalence of infection varied markedly between the centers (28–74%)[26].

Needing a caregiver was associated with an increased risk of mortality and with DFU recurrence, while it was negatively associated with ulcer healing. The need for a caregiver may be regarded as a surrogate of frailty. It is well known that frailty is a clinical syndrome associated with dependence and mortality in the older adults, including those with diabetes. Moreover, frailty may be a more powerful prognostic marker than the burden of comorbidity itself [36,37,38]. This was also the case in the study by Gershater et al. [7], where the authors analyzed a cohort of 2,480 diabetics with a first ulcer and observed that patients with an informal caregiver patient were more likely to have a major amputation or to have died before healing compared to those who did not have a caregiver (odds ratio (OR): 2.16, 95% CI 0.43 – 3.28. p < 0.005). The role of informal caregivers remains largely unexplored, and its importance is fundamental in the care of a patient with an ulcer [39].

There are many classifications of people with diabetes mellitus and DFU [14]. The PEDIS classification was designed as an aid for prospective research [14]. Using this classification, Chuan et al. [18] created a scoring system with 364 diabetic foot patients treated in a hospital with a mean follow up of 25 months. The study outcomes were healed DFU and a combination of unhealed DFU, amputation and death. They observed that a PEDIS classification score with a value of at least seven was associated with the worst clinical prognosis of the patients. In our study, a PEDIS score of ≥ 7 was associated with an increased risk of ulcer recurrence during the follow-up period.

Ince et al. [17] conducted a study with diabetic patients with foot ulcers from the UK, Germany, Tanzania and Pakistan to determine the prognostic value of the SINBAD classification score for healing vs no healing, including amputation or death. The authors observed that despite all the differences between countries, ulcers with a SINBAD score of at least 3 had a worse clinical prognosis. In the annual report of the UK National Diabetes Foot Care Audit [40] from 2018, with 19,453 patients with DFU, the SINBAD classification was also used for the same purpose. It was observed that patients with a SINBAD score equal to or greater than 3 were less likely to be alive and ulcer-free at 12 and 24 months. Our study observed that a SINBAD score ≥ 3 was associated with a risk of non-healing during the 12 months follow up period.

This study has some limitations. We have no information on follow-up in 34 patients who discontinued the study. Some limitations, such as possible underreporting, selection bias, and the absence of socioeconomic data, as well as the absence of the prevalence of mental health disorders (depression and anxiety) were previously acknowledged in a prior study [15].