Estimating the diagnostic accuracy of the ankle-brachial pressure index for detecting peripheral arterial disease in people with diabetes: A systematic review and meta-analysis

AIM

To systematically evaluate research investigating the accuracy of the ankle-brachial index (ABI) for diagnosing peripheral artery disease (PAD) in people with diabetes, as the accuracy is thought to be reduced in this cohort.

METHODS

A database search of EBSCO Megafile Premier, Embase and The Cochrane Library was conducted to 28 February 2019. Prospective and retrospective investigations of the diagnostic accuracy of the ABI for PAD in people with diabetes using an imaging reference standard were eligible. Sensitivity and specify of the ABI and bivariate meta-analysis against reference tests, or a standard summary receiver operating curve analysis (SROC) was performed.

RESULTS

Thirty-three studies met the inclusion criteria. ABI was compared with angiography in 12 studies and with colour duplex ultrasound (CDUS) in 21 studies. A SROC analysis of studies using angiography as the reference standard found a diagnostic odds ratio (DOR) of 9.06 [95% confidence interval (CI) 3.61 to 22.69], and area under the curve (AUC) of 0.76 (95% CI 0.66 to 0.86). Bivariate analysis of studies using CDUS demonstrated mean sensitivity of 0.60 (95% CI 0.48 to 0.71; P = 0.097) and mean specificity of 0.87 (95% CI 0.78 to 0.92; P < 0.001) with a DOR of 9.76 (95% CI 5.24 to 18.20; P < 0.0001) and AUC 0.72.

CONCLUSIONS

These results suggest the ABI has a high specificity but lower sensitivity in detecting imaging diagnosed PAD in people with diabetes. The low probability of the testing being able to rule diagnosis in or out suggest that the ABI has limited effectiveness for early detection of PAD in this cohort.

Prevalence of ankle equinus and correlation with foot plantar pressures in people with diabetes

BACKGROUND

An association between equinus and plantar pressure may be important for people with diabetes, as elevated plantar pressure has been linked with foot ulcer development. To determine the prevalence of equinus in community dwelling people with diabetes and to examine any association between presence of equinus and forefoot plantar pressures.

METHODS

Barefoot (Tekscan HR Mat™) and in-shoe (Novel Pedar-X®) plantar pressure variables, non-weight bearing ankle range of motion and neuropathy status were assessed in 136 adults with diabetes (52.2% male; 47.8% with neuropathy; mean (standard deviation) age and diabetes duration: 68.4 (11.5) and 14.6 (11.1) years respectively).

FINDINGS

Equinus, when measured as ≤5° dorsiflexion, was present in 66.9% of the cohort. There was a significant correlation between an equinus and barefoot (r = 0.247, p = 0.004) and in-shoe forefoot pressure time integrals (r = 0.214, p = 0.012) and in-shoe forefoot alternate pressure time integrals (r = 0.246, p = 0.004). Significantly more males (p < 0.01) and people with neuropathy (p = 0.02) or higher glycated haemoglobin levels (p < 0.01) presented with an equinus.

INTERPRETATION

Community dwelling adults with diabetes have a high rate of ankle equinus which is associated with increased forefoot pressure time integrals and a two-fold increased risk of high in-shoe peak pressures. Clinical assessment of an ankle equinus may be a useful screening tool to identify adults at increased risk of diabetic foot complications.

Validation of a weight bearing ankle equinus value in older adults with diabetes

Background

Accurate measurement of ankle joint dorsiflexion is clinically important as a restriction has been linked to many foot and ankle pathologies, as well as increased ulcer risk and delayed ulcer healing in people with diabetes. Use of the reliable weight bearing (WB) Lunge test is limited as normal and restricted ranges for WB ankle joint dorsiflexion are not identified. Additionally the extent of WB dorsiflexion restriction that results in clinically adverse outcomes is unclear. Therefore the aim of this investigation is to validate a proposed weight bearing equinus value (dorsiflexion < 30°) in unimpaired cohorts, and secondly to investigate any clinical effects this degree of ankle dorsiflexion restriction has on forefoot plantar pressure variables in older adults with diabetes.

Methods

Ankle dorsiflexion was measured using a Lunge test with the knee extended in young adults without diabetes (YA) and older adults with diabetes (DA). In-shoe and barefoot plantar pressure was recorded for the DA group. Spearman’s correlation was calculated to determine any association between the presence of ankle equinus and plantar pressure variables in the DA group. DA group differences in people with and without an equinus were examined.

Results

A weight bearing equinus of < 30°, assessed in a lunge using an inclinometer placed on the anterior tibia, falls within the restricted range in young unimpaired cohorts. In the DA group this degree of ankle restriction had a fair and significant association with elevated barefoot forefoot peak pressure (r = 0.274, p = 0.005) and pressure-time integrals (r = 0.321, p = .001). The DA equinus group had significantly higher barefoot peak pressure (mean kPa (SD): 787.1 (246.7) vs 652.0 (304.5), p = 0.025) and pressure-time integrals (mean kPa (SD): 97.8 (41.6) vs 80.4 (30.5), p = 0.017) than the DA non equinus group.

Conclusions

We support a preliminary weight bearing ankle equinus value of < 30°. This value represents a restricted range in young adults and is correlated with increased forefoot plantar pressure variables in older adults with diabetes. Mean population weight bearing ankle dorsiflexion data presented here for older adults with diabetes, will allow use of the more functional Lunge test with knee extended in research and clinical practice.

Weight bearing versus non-weight bearing ankle dorsiflexion measurement in people with diabetes: a cross sectional study

Background

Accurate measurement of ankle dorsiflexion is important in both research and clinical practice as restricted motion has been associated with many foot pathologies and increased risk of ulcer in people with diabetes. This study aimed to determine the level of association between non-weight bearing versus weight bearing ankle dorsiflexion in adults with and without diabetes, and to evaluate the reliability of the measurement tools.

Methods

One hundred and thirty-six adults with diabetes and 30 adults without diabetes underwent ankle dorsiflexion measurement non-weight bearing, using a modified Lidcombe template, and weight bearing, using a Lunge test. Pearson product-moment correlation coefficients, intraclass correlation coefficients (ICCs) with 95% confidence intervals, standard error of measurement and minimal detectable change were determined.

Results

There was a moderate correlation (r = 0.62–0.67) between weight and non-weight bearing tests in the non-diabetes group, and a negligible correlation in the diabetes group(r = 0.004–0.007). Intratester reliability was excellent in both groups for the modified Lidcombe template (ICC = 0.89–0.94) and a Lunge test (ICC = 0.83–0.89). Intertester reliability was also excellent in both groups for the Lidcombe template (ICC = 0.91) and a Lunge test (ICC = 0.88–0.93).

Conclusions

We found the modified Lidcombe template and a Lunge test to be reliable tests to measure non-weight bearing and weight bearing ankle dorsiflexion in adults with and without diabetes. While both methods are reliable, further definition of weight bearing ankle dorsiflexion normative ranges may be more relevant for clinical practice.

Association between ankle equinus and plantar pressures in people with diabetes. A systematic review and meta-analysis

BACKGROUND

Diabetes is one of the most common chronic diseases in the world and is associated with a life-time risk of foot ulcer of 12-25%. Diabetes related restriction in ankle joint range of dorsiflexion is proposed to contribute to elevated plantar pressures implicated in the development of foot ulcers.

METHODS

A systematic search of EBSCO Megafile Premier (containing MEDLINE, CINAHL, SPORTSdiscus and Academic Search Complete) and The Cochrane Library was conducted to 23rd November 2016. Two authors independently reviewed and selected relevant studies. Meta-analysis of study data were conducted where possible.

FINDINGS

Fifteen studies met the inclusion criteria. Three studies were eligible to be included in the meta-analysis which found that equinus has a significant, but small, effect on increased plantar pressures (ES=0.26, CI 95% 0.11 to 0.41, p=0.001). Of the remaining studies, eight found evidence of an association between limited ankle dorsiflexion and increased plantar pressures while four studies found no relationship.

INTERPRETATION

Limited ankle joint dorsiflexion may be an important factor in elevating plantar pressures, independent of neuropathy. Limited ankle dorsiflexion and increased plantar pressures were found in all the studies where the sample population had a history of neuropathic foot ulceration. In contrast, the same association was not found in those studies where the population had neuropathy and no history of foot ulcer. Routine screening for limited ankle dorsiflexion range of motion in the diabetic population would allow for early provision of conservative treatment options to reduce plantar pressures and lessen ulcer risk.

Exercise and ectopic fat in type 2 diabetes: A systematic review and meta-analysis

Ectopic adipose tissue surrounding the intra-abdominal organs (visceral fat) and located in the liver, heart, pancreas and muscle, is linked to cardio-metabolic complications commonly experienced in type 2 diabetes. A systematic review and meta-analysis was performed to determine the effect of exercise on ectopic fat in adults with type 2 diabetes. Relevant databases were searched to February 2016. Included were randomised controlled studies, which implemented≥4 weeks of aerobic and/or resistance exercise and quantified ectopic fat via magnetic resonance imaging, computed tomography, proton magnetic resonance spectroscopy or muscle biopsy before and after intervention. Risk of bias and study quality was assessed using Egger's funnel plot test and modified Downs and Black checklist, respectively. Of the 10,750 studies retrieved, 24 were included involving 1383 participants. No studies were found assessing the interaction between exercise and cardiac or pancreas fat. One study assessed the effect of exercise on intramyocellular triglyceride concentration. There was a significant pooled effect size for the meta-analysis comparing exercise vs. control on visceral adiposity (ES=-0.21, 95% CI: -0.37 to -0.05; P=0.010) and a near-significant pooled effect size for liver steatosis reduction with exercise (ES=-0.28, 95% CI: -0.57 to 0.01; P=0.054). Aerobic exercise (ES=-0.23, 95% CI: -0.44 to -0.03; P=0.025) but not resistance training exercise (ES=-0.13, 95% CI: -0.37 to 0.12; P=0.307) was effective for reducing visceral fat in overweight/obese adults with type 2 diabetes. These data suggest that exercise effectively reduces visceral and perhaps liver adipose tissue and that aerobic exercise should be a key feature of exercise programs aimed at reducing visceral fat in obesity-related type 2 diabetes. Further studies are required to assess the relative efficacy of exercise modality on liver fat reduction and the effect of exercise on pancreas, heart, and intramyocellular fat in type 2 diabetes and to clarify the effect of exercise on ectopic fat independent of weight loss.

The effect of diabetic neuropathy on foot bones: a systematic review and meta-analysis

AIMS

It is proposed that diabetic neuropathy may affect peripheral bone. Direct innervation of bone as well as neural control over its vascular supply and muscular influences may be affected by diabetes-induced peripheral neuropathies. Associated changes to bone may contribute to the occurrence of foot bone pathology in this population. This systematic review aims to examine the literature related to the effect of diabetic neuropathy on foot bones.

METHODS

Studies examining relationships between neuropathy and indicators of bone health (e.g. bone mineral density) in populations with diabetes were sought. Relevant publications were obtained from searches in MEDLINE, CINAHL and Embase in the period up to March 2013. Meta-analysis was performed using a random effects model in the statistical package Stata version 12.1.

RESULTS

Ten studies met the inclusion criteria and were included in the narrative synthesis. All studies were cross-sectional or case-control in design. Four of the 10 included studies found results indicating poorer bone health in those with diabetes and neuropathy compared with those with diabetes without neuropathy. Seven of the 10 studies were able to be included in a meta-analysis. The mean pooled effect was -0.36 (95% CI -0.76 to 0.04; P = 0.08), indicating a non-significant trend towards poorer bone health in those with diabetic neuropathy.

CONCLUSIONS

We did not find a significant relationship between presence of neuropathy in those with diabetes and poorer peripheral bone health. However, methodological limitations of the included studies mean further research is required to investigate this theoretical relationship.