Baseline characteristics and risk factors for ulcer, amputation and severe neuropathy in diabetic foot at risk: the BRAZUPA study

Categorical data is shown as percentages; continuous variables as means (standard deviation—SD).

Comparison between groups of categorical data was made using the Chi Square test. Differences in continuous variables between two groups were evaluated by Student’s t test.

Independent risk factors for ulcer and amputation were assessed by multiple logistic regression analysis. The model for ulcer included age, gender, type of diabetes, disease duration, ethnicity, BMI, smoking status, foot at risk classification (foot without risk as reference category), vibratory sensation, visual impairment, hypertension, renal impairment and macro region of origin. Adjustments were done by the forward conditional method. The model for amputation included all of the abovementioned variables with the addition of history of previous ulcer episode.

Statistical analysis was performed with SPSS version 20.0 (IBM Inc.). Significance was defined as a p value <0.05.

Baseline patient characteristics are summarized in Table 1. The mean age and disease duration of participants was 57.67 (14.23) years and 14.21 (9.77) years, respectively. The majority of patients were female (58.6 %), White (53.8 %) and had type 2 diabetes (90.7 %). Nearly a quarter (25.3 %) of patients had a previous history of ulcer and 13.7 % of amputation (17.3 % of these were major amputations). One in every twenty patients (5.3 %) had undergone more than one amputation procedure. One-third of patients had neuropathic foot (33.7 %) and only 8.5 % had ischemic disease alone. Nearly one-fifth (18.6 %) of all evaluated patients presented a current, active foot ulcer at evaluation.

In the analysis of patients with moderate/severe vs. normal/mild neuropathic symptom scores, those with higher (more severe) scores were more frequently female (53 vs. 45.8 %; p = 0.01), older (58.5 ± 13 vs. 56.4 ± 15.5 years; p = 0.009) and had a higher BMI (29.95 ± 6.58 vs. 28.92 ± 5.79; p = 0.004). White patients were less frequently classified as having severe symptoms than those of Black, Mixed or Asian ethnicities (45.8 vs. 52.9, 54.5 and 57 %, respectively; p = 0.02). Patients with retinopathy had a higher prevalence of severe symptoms (52.7 vs. 45.8 %; p = 0.02), as did those with visual impairment (54.1 vs. 46.1 %; p = 0.006). In contrast to what was found for amputation and ulcer, patients from the emerging regions had a higher prevalence of severe symptoms than those coming from developed regions (57.3 vs. 44.9 %, respectively; p < 0.001).

Table 4 shows the prevalence of clinical characteristics in patients with normal/mild and moderate/severe symptoms.

Multivariate regression analyses are summarized in Table 5. Independent risk factors for ulcer were male gender, smoking, foot at risk classification (highest risk associated with the neuroischemic group and lowest to the ischemic group), region of origin (higher risk for those from more economically developed regions), presence of retinopathy and absent vibratory sensation. Risk factors for amputation were male gender, type 2 diabetes, foot at risk classification (higher risk for ischemic foot), hypertension (lower risk), region of origin (South/Southeast), previous history of ulcer and altered vibratory sensation. There was no association between the two outcomes and ethnicity in multivariate analyses.

This is the first large, comprehensive multicenter epidemiological study of foot at risk in diabetes outside Europe [9], Australia [11] or China [12]. We found the prevalence of previous ulcer to be 25 and 14 % for amputation. Additionally, 18 % of patients had an active ulcer at the time of evaluation. Although at a first glance the rate of active foot ulcers could be considered high when compared to the results of other studies (e.g. the FREEMANTLE Study) [11], there are some important differences. Our results are, at this time, from a transversal study and not prospectively evaluated. Additionally, the inclusion of centers with structured foot care units could have contributed to increase active ulcer prevalence. In the next years, in a prospective basis it will be possible to determine with more certainty if the prevalence of ulcers in patients with diabetes in Brazil is higher than in other parts of the world.

Fifty percent had moderate/severe neuropathic symptoms. These findings are in line with previous data [13, 14], especially that from developing nations [15, 16], except for the high proportion of patients with severe symptoms. Also noteworthy, despite the high number of patients with symptoms, we found that only 14 % were receiving medication for neuropathic pain relief.

Our patients had a unique combination of characteristics. The major difference between our patients and those evaluated in studies from the USA [17] or Western Europe [18] is a higher proportion of patients with neuropathic disease, with clear predominance of the neuroischemic population, and a smaller number of patients with isolated ischemic disease. Most studies conducted in economically developed nations show a high prevalence of ischemic disease [19, 20]. Similar findings had been found in one of the centers involved in our study, but with a higher proportion of isolated neuropathic disease [21]. Mean age (57 years old) was also younger than that reported for patients from Western Europe [22] and North America [23]. On the other hand, characteristics like longer disease duration and high frequency of associated comorbidities are in line with data from developed nations [18‐23].

Individuals with more severe symptoms were more commonly overweight, older and female. The relation between BMI and symptoms of neuropathy could be associated with other obesity related comorbidities or with excessive load on the lower extremities. Ulcer and amputation were less common in patients of Asian descent, a fact already reported in other studies [24]. However, in multiple adjusted regressions, ethnicity was not a factor associated with either outcome. We believe that this is due to the historic and strong miscegenation typical of our country’s population. Social and economic factors are probably more associated with these ethnic differences in terms of the prevalence of ulcer and amputation [25].

Indeed, ulcer and amputation were more common in the more developed regions of the country, which may be related to the greater prevalence of other diseases such as obesity and metabolic syndrome [26]. Also, social and economic differences might play an important role, in conjunction with disparities in health care access. Increased awareness by health care professionals and more swiftly available access to specialty centers could result in increased diagnosis, which in turn is reflected in an increase in apparent prevalence. Additionally, diabetes prevalence is higher in the South and Southeast regions of the country [27].

Regarding factors independently associated with ulcer risk, the finding that men have an increased risk of ulcer is in accord with the great majority of previous studies [9, 11, 12, 21, 22, 25]. This fact is associated with disease progression and also with social facts related to men being more frequently family providers and consequently showing lower adherence to medical visits and wound care [28].

Loss of vibratory sensation and large fiber damage are frequently seen as the first sensations affected by distal diabetic polineuropathy [29]. Retinopathy is also a long-term complication of diabetes and is expected to co-exist with neuropathy. In addition, it can make self-care and self-examination more difficult and thus predispose to ulcer. In adjusted analysis, however, visual impairment was not an associated risk factor, indicating that microvascular processes occurring simultaneously is a more likely explanation for ulcer than visual deficits.

The lowest risk for ulcer relative to foot at risk classification when compared to foot without risk was isolated arterial disease. The greatest risk was linked to neuroischemic disease. The combination of neuropathy and peripheral arterial disease has been associated with a high ulcer risk in most previous studies [9, 11, 21], but ischemia is normally associated with a greater risk than neuropathy [9, 30]. This difference might be due to neuropathy being more common in developing countries. Similar findings have been published in smaller studies from other emerging nations [12, 15, 16, 21]. A younger age at presentation and a lower prevalence of obesity and other associated diseases apart from diabetes itself may combine to expose patients to ulcer sooner than atherosclerotic disease manifests.

As reported for ulcer, male sex is associated with a greater risk of amputation in almost all studies on diabetic foot complications. In the case of type 2 diabetes, it is more commonly associated with other comorbidities that can predispose to poorer outcomes in ulcer, such as obesity, insulin resistance, dyslipidemia and atherosclerosis [31], thus explaining the greater risk of amputation, even in a younger population. Loss of vibratory sensation and previous ulcer are established risk factors for amputation. Amputation is, in fact, a last resort treatment for diabetic ulcers.

Unlike ulcer, patients presenting ischemic disease, either in an isolated form or associated with neuropathy presented the greatest risk for amputation in multivariate analysis. Ischemic disease has frequently been associated with poorer ulcer outcomes [9, 11], especially when associated with infection [32], explaining the higher probability of ending in amputations. The finding of hypertension being associated with a lower risk for amputation even in adjusted analysis was a surprising one, and no clear explanation could be found for this fact with the current data. Follow up results may cast some light on this matter.

This study has some limitations. The data represent baseline information, and the causative factors here identified have the constrictions inherent to cross sectional evaluations. However, all the risk factors found for ulcer and amputation are well recognized in the literature, which makes us more confident that our estimates are adequate. Another limitation is that we found a relatively high number of patients referred for evaluation who in the end were revealed to present foot without risk. Although serving as a reference category to estimate the risk of neuropathy vs. ischemia for ulcer and amputation, the relatively small number of patients with foot at risk could have underpowered our analysis. Nevertheless, we were able to identify several risk factors associated with those conditions, a detail that reassures us that the analysis deserves merit. Additionally, our population was a large one, and the fact that patients came from several different centers and were primary care referrals support the representativeness of our sample. As with any other studies using ABI for ischemic disease diagnosis, the prevalence of ischemic disease could have been underrepresented, but this methodology has been used in most studies; therefore, data from this study is comparable to that in others. Whether our findings can be extrapolated to other South or Latin American populations is also debatable as our country bears significant ethnic and social differences from others in this region of the world.