Epidemiology of diabetic foot disease and etiology of ulcerationChapter 1

Diabetic foot ulceration represents a major medical, social, and economic problem all over the world. Although >5% of patients with diabetes have a history of foot ulceration, the cumulative lifetime incidence may be as high as 25%. Foot problems in patients with diabetes account for more hospital admissions than any other long-term complications of diabetes and also result in increasing morbidity and mortality.

As reviewed in this chapter, peripheral neuropathy, arterial disease, and foot deformities are the main factors accounting for this increased risk. Age and sex as well as social and cultural status are contributing factors. Knowing these factors is essential to classify every individual with diabetes using a risk grading system and to take preventive measures accordingly.

The diabetic foot defined as infection, ulceration, and/or destruction of deep tissues associated with neurological abnormalities and various degrees of peripheral arterial disease in the lower limb presents a particularly troubling picture and it has been claimed that every 20 seconds, a lower limb is amputated due to diabetes (Int Work Group 2011). It is estimated that 50–70% of all lower extremity amputations are related to diabetes. Of all amputations in patients with diabetes, 85% patients are preceded by a foot ulcer that subsequently deteriorates to a severe infection or gangrene requiring amputation. Four out of five ulcers are claimed to be preceded by external trauma. Diabetic foot complications result in huge costs for both society and people living with diabetes. Foot problems use 12–15% of the health-care resources for diabetes. In developing countries figures up to 40% have been reported (Boulton 2005, Driver et al. 2010).

The major adverse outcomes of diabetic foot problems are foot ulcers and amputations. Up to 85% of all amputations begin with an ulcer; every year, approximately 4 million more people develop a diabetic foot ulcer. More information is available on numbers of amputations than on numbers of ulcers. Until proper population-based registers of people with diabetes are available, reliable data relating to accurate estimates of the prevalence and incidence of these late complications will be limited to community-based studies or studies from dedicated centers or clinics.

As most of the information in the current scientific literature comes from selected populations, and different definitions are used, it is difficult to gauge the extent of foot problems worldwide. It is also likely that the type of ulcer varies around the world: in developed countries, up to 60% of new ulcers are associated with peripheral arterial disease so called neuroischemic and ischemic ulcers; in developing countries, neuropathic ulcers of various origins are more common.

The point prevalence of foot ulcers in developed countries varies between 1.5% and 10% in various populations; a corresponding incidence of 2.2–5.9% has been reported from studies in Western Europe and North America. Foot ulcers occur in both type 1 diabetes and type 2 diabetes. In elderly patients with diabetes type 2 the reported prevalence of foot ulcers has been 5–10%. In studies that focused on younger subjects with type 2 diabetes or individuals with diabetes type 1, the estimated prevalence was 1.7–3.3% (Int Consensus 2011, Richard et al. 2008, Rathur & Boulton 2007, Reiber & LeMaster 2008). In community-based European studies, the prevalence varies between 1.4% and 8.3% and in clinic-based studies from developing countries, a prevalence of 3.6–11.9% is reported especially from Arab countries with a prevalence of diabetes of 19.2–29.2%. In western countries, on 4average 2 out of 100 individuals with diabetes have a foot ulcer (Richard et al. 2008). The prevalence increases in populations in the presence of predisposing factors.

Two studies from Northern European countries reported the annual incidence of foot ulcers in the general population to be just >2% (Abbot et al. 2002, Muller et al. 2002, Richard et al. 2008). However, incidences of 2.5–7.2% have been reported in high risk or selected populations compared with 0.6–2.2% in European community-based studies and 3–6% in clinic-based studies from developing countries (Abbas et al. 2002, Kumar et al. 1994, Ramachandran 2004, Richard et al. 2008). Ulceration is much more common in patients with predisposing risk factors; annual incidence rates in neuropathic individuals vary from 5% to over 7% (Abbot et al. 2002, Manes et al. 2002). It is likely that the cumulative lifetime incidence of foot ulcers may be as high as 25% or even higher in individuals with diabetes, and a previously healed ulcer 30–50% will have a new foot ulcer within 2 years (Abbot et al. 2002, Boulton et al. 2005, Rathur & Boulton 2007).

Studies from the United Kingdom suggest that foot ulcers and amputation are less common in Asian patients of Indian subcontinent origin (Abbot et al. 2005). Possible explanations for the findings in Asian patients relate to differences in limited joint mobility and to better foot care in certain religious groups such as Muslims. North American studies reported that ulceration was more common in Hispanic Americans and in Native Americans than in non-Hispanic whites (Lavery et al. 2003, Resnick et al. 1999). Foot ulceration also appears to be associated with social deprivation (Reiber & LeMaster 2008). In most publications on foot ulcers, there is a male dominance (Gershater 2008, Prompers et al. 2008).

In many reports, prevalence data rely on hospital discharge or claims data (Driver et al. 2010, Payne & Scott 1998, Rathur & Boulton 2007, Reiber & LeMaster 2008). The rate of hospital discharges for patients with diabetes in the United States with leg/foot ulcers for every 1000 patients with diabetes rose from 5.4 in 1980 to 6.9 in 2003. Ulcer prevalence among persons aged >44 years was 6.5/1000 patients with diabetes and it rose progressively to 10.3/ 1000 in individuals aged >75 years (Reiber & LeMaster 2008).

Adequately performed population-based studies regarding the incidence and prevalence of foot ulcer in the lower leg are scarce. Most studies are performed from the perspective of patients attending a dedicated clinic or in a selected study population. The difference in incidence results, in many cases, from differences in the design of the study, demographic factors, and the prevalence of diabetes, as well as variations in registration systems. A major issue is that a foot ulcer is frequently not reported/detected by the patient himself. In one study, 25% of the patients with a full skin foot ulcer denied the presence of the ulcer at an interview followed by a physical examination (Apelqvist & Larsson 2000). In 10% of patients with an ulcer below the ankle, the diagnosis of diabetes was previously unknown. In 50% of patients with diabetes mellitus and a foot ulcer, the treating physician was not aware of the ulcer and the ulcer was not recorded in the medical records. As a consequence, studies tend to underestimate the total number of foot ulcers. The study design and method are essential in that respect. If it is a survey with self-reported questionnaire, interview, analysis of medical records, or cross-sectional clinical study, this will substantially influence the outcome (Apelqvist & Larsson 2000, Reiber & LeMaster 2008). In most countries, the incidence of foot ulcers is underestimated as a continuous registration system is not in place.

Sensory neuropathy is associated with the loss of pain, awareness of pressure, temperature sensation, and proprioception. Due to the lack of protective sensation, the foot is vulnerable to unattended minor injuries caused by excess pressure or mechanical or thermal injury. Thus, acute injury, ill-fitting shoes, or walking barefoot can precipitate an ulcer. As a consequence, a minor trauma that is not felt by the patient can quickly cause skin damage or ulceration. Prospective studies have shown that sensory neuropathy is a major predictor for the development of foot ulcers.

Motor neuropathy, affecting both the intrinsic foot muscles and leg muscles, results in atrophy and weakness of the intrinsic muscles of the foot, flexion deformities of the toes, and an abnormal walking 5pattern that alters the biomechanics and gradually the foot anatomy due to postural instability and disturbed coordination caused by the lack of proprioceptive feedback from the lower extremities.

Autonomic neuropathy results in reduced or absent sweat secretion, causing dry skin, cracks, and fissures. Foot deformities, abnormal gait, and limited ‘joint mobility syndrome’ create a set of conditions that result in altered biomechanical loading of the foot, with increased plantar foot pressures and increased shear forces. The ‘stiff foot syndrome’ with impaired joint mobility is probably caused by disturbed collagen metabolism, or nonenzymatic glycosylation of proteins in the joint tissue and skin.

Foot deformities, limited joint mobility, and altered loading of the foot are obvious consequences arising from these pathological processes. These lead to excess pressure, skin thickening, and callus formation, which again increase the already abnormal loading and, often, subcutaneous hemorrhage and ulcer. The neuropathic bone and joint disease, usually affecting mid or hindfoot (sometimes referred to as ‘Charcot foot’ or neuro-osteoarthropathy) can cause additional severe deformity and plantar ulcer.

Ulcers frequently result from external trauma to the insensitive foot, such as ill-fitting shoes, burns, walking bare foot, and foreign objects in shoes (Apelqvist & Larsson 2000, Gershater 2008). An ulcer caused by increased mechanical stress is usually localized to the metatarsal heads or plantar of the first digit, whereas decubitus ulcers are located to the heel. An ulcer caused by ill-fitting shoes or acute trauma is usually localized to the digits or dorsum of the foot. Ischemic ulcers are usually located to the tips of toes and the lateral part of the foot. Interdigital ulcers are usually caused by skin maceration and cracked skin. The distribution of type, site, and cause of ulcer country-by-country depends on the climate, access to and use of footwear, physical activity, and culture (Abbas & Archibald 2005, Int Work Group 2011).

With regard to the etiology of foot ulceration, peripheral arterial disease and neuropathy are frequently present in the same patient. Although it is traditionally stated that the majority of foot ulcers are purely neuropathic, an increase has been found in the incidence of neuroischemic and/or ischemic foot ulcers. In large cohort studies in Europe, neuroischemic or ischemic ulcers accounted for 50–58% of all diabetic foot ulcers admitted to specialist care (Gershater 2008, Prompers et al. 2008).

There are no peripheral artery lesions specific to diabetes, but the pattern of atherosclerosis is different, typically affecting more distal segments, and is of greater prevalence than in those without diabetes (20–40%) (Int Work Group 2011). Peripheral arterial disease, usually in conjunction with minor trauma, may result in painful, purely ischemic foot ulcers. However, peripheral arterial disease and neuropathy are frequently present in the same patient. An important observation relating to the consequence of peripheral arterial disease is that many individuals with diabetes ischemia do not have severe rest pain or claudication. This is believed to be related to the loss of sensation due to peripheral sensory neuropathy.

It is essential to distinguish between different types of ulcers, especially with respect to predisposing factors such as neuropathy or neuroischemia/ischemia. Plantar neuropathic foot ulcers constitute 18–23% of foot ulcers in large European cohort studies (Gershater 2008, Prompers et al. 2008), but are substantially more frequent in countries of the Middle East and North Africa (Abbas et al. 2002, Ramachandran 2004, Zhangrong et al. 1997). It should also be recognized that in many cases, the ulcer is a sign of an underlying multiorgan disease (cardio/cerebrovascular disease and nephropathy) (Gershater 2008, Prompers et al. 2008). Foot ulcers vary in the range and extent of tissue damage and wound characteristics. A classification and grading system are essential both for management and documentation and many have been suggested. The PEDIS system has been developed to describe the complexity of the diabetic foot ulcer (Peters & Lavery 2001).

In most studies, irrespective of design and study population and excluding amputations due to trauma and tumor, the annual incidence ranges from 46.1 to 936 per 100,000 people with diabetes (Larsson & Apelqvist 1995, Reiber & LeMaster 2008). According to the Diabetes Atlas, the middle 50% of amputation prevalence ranged between 0.9% and 2.4%, with a corresponding incidence of amputation of between 181 and 463 per 100,000 people with diabetes.

Most of these data are collected from Western Europe and United States, but there is an increasing concern regarding the development in other parts of the world especially the Pacific, Middle East, and Northern Africa (LEA study 2000), with the prevalence of diabetes-related amputation at around 0.7–5%. There are substantial differences both with regard to incidence and prevalence of diabetes-related amputations globally.

However, for these comparisons to be valid there are many possible pitfalls that need to be avoided (van Houtum & Lavery 1997). A significant number of epidemiological and clinical considerations are to be dealt with and many of these are extremely important to realize when comparing ulceration or amputation rates. The influence of using a certain definition of numerators and denominators is explained. Also, the clinical setting in which the research has been performed is of importance. Many possible pitfalls are discussed in this presentation. Some of these potential pitfalls are listed below. A variety of numerators and denominators have been used in published reports on diabetes-related amputations.

Although amputation appears to be a clear defined end point, there are several other methodological issues. The difficulty in evaluating epidemiological data with regard to incidence of amputation is well recognized (Jeffcoate & van Houtum 2004, Larsson et al. 2008, Schaper et al. 2012). There are many important confounders explaining differences in data collection and reflecting the complexity of data involved in any detailed presentation of amputation epidemiology. Adequately performed population-based studies regarding the incidence of amputations in the lower leg are scarce.

A fundamental methodological question pertaining to all epidemiological studies is the completeness of the material with regard to diabetes. Most amputation studies are based on central registers and subject to any shortfall inherent in such registers. The shortcomings of national registers are well known (Larsson & Apelqvist 1995, Schaper et al. 2012). A major issue is that the diagnosis of diabetes is not established or not recorded at the time of amputation.

The diagnosis of diabetes is frequently defined as previously known diabetes, usually identified as treatment with insulin or oral agents. In a study by Leslie, in 19% of diabetes-related amputations, the diagnosis of diabetes was only established/recognized at the time of amputation (Valeri et al. 2004). Corresponding underestimations have been reported from other areas. Moreover, in most studies, amputations are lumped together but a single toe, a whole forefoot, or an above-knee amputation has a very different impact on the patient.

Not only does a patient have two legs but, taking minor amputations into consideration, also all ten toes are potential subjects for amputation, simultaneously or on different occasions. Central registers usually differentiate neither left from right side nor primary amputation from reamputation. After healing of one or more minor amputations, a major amputation on the same side may follow, perhaps years later, and thus several final amputation levels have to be considered in the same patient. Amputations may also have taken place before the time period of the study. All these are events that may influence incidence calculations. However, to reflect a true epidemiological picture, final amputation levels must have preference over primary amputation levels. Clear definitions with regard to all these aspects are mandatory to make comparisons between studies possible.

Another issue is whether the study focuses on a geographical area or is hospital/clinic-based. The setting as well as a referral (primary/secondary) or tertiary unit is also of importance. Amputations may be performed in a clinic or hospital that is not the site of the clinical study (and therefore may go unrecognized). The health-care structure and reimbursement routines have been reported to influence the reported amputation rate. All routine registers have a certain inherent shortfall. By systematic scrutiny of different independent registration systems, these pitfalls can be reduced (this includes data/records from both in- and outpatient clinics involved in treating patients with diabetes and operation theatre records) (Larsson et al. 2008).7

Ideally when evaluating diabetes-related amputations, the total amputation rate irrespective of level should be reported as well as indications and rational for level selection.

Several population-based studies observed a significant reduction in major amputations over time (Krishnan et al. 2008), and after correction for the increasing number of people with diabetes, in some countries a relative decrease was observed over a longer period of time in the number of lower extremity amputations in people with diabetes. However, there are also several countries that report an increase in rates of amputation. The reason for this discrepancy is unclear, but factors such as health-care organization and reimbursement might be, in part, responsible (Apelqvist & Larsson 2000, Reiber & LeMaster 2008).

When assessing the impact of an interventional diabetic foot program, a number of basic underlying factors, such as prevalence of diabetes, the age profile of the population, comorbidity, and smoking habits, must be considered. Such factors may mask the effect of intervention unless compared with a situation where such intervention is not applied and must be considered when assessing future incidence rates (Schaper et al. 2012).

The differences in incidence results, in many cases, are established as a result of differences in the design of the study, demographic factors, and the prevalence of diabetes, as well as variations in registration systems, and differences in reimbursement for a variety of procedures, e.g. minor foot debridements or amputations and revascularizations (Holman et al. 2012, Vamos et al. 2010).

Leg amputations are associated with an increase in mortality in people with diabetes. By the time an amputation is necessary, people have usually had diabetes for many years, and often have severe comorbidity. Death around the time of the amputation (perioperative mortality) occurs in up to 10% of cases. Mortality rates increase over the 5 years following amputation: 30% of patients die within 1 year, 50% die within 3 years, and 70% die within 5 years (Larsson et al. 1998). 8In developing countries, these figures tend to be even higher because many people seek medical attention only when their foot problem is so far advanced that their limbs and their lives appear immediately threatened.

Amputation rate should therefore not be used as a sole quality indicator in diabetic foot disease, unless it can be corrected for the relevant characteristics of the patient, the leg and the foot (Schaper et al. 2012). Although there are limitations to clinical research using amputation as an end point, it seems likely to remain the most commonly used way to compare health strategies and to determine the importance of an intervention. The incidence and prevalence figures are helpful in assessing any potential beneficial effect to justify an intervention if they are used and interpreted with caution. Given the limitations explained above, epidemiological research concerning the prevalence and incidence of foot-related complications still forms the backbone of clinical research in the area of the diabetic foot.