This was a cross-sectional observational, single-center, descriptive and analytical study. It was conducted over a six-month period from November 1, 2024, to April 30, 2025, at the outpatient consultation of the endocrinology and diabetology department of Yopougon Atiié General Hospital in Abidjan. We included in this study diabetic patients over 18 years of age, regardless of the type of diabetes. Patients who refused to participate as well as non-autonomous patients were excluded from the study.

An information form on epidemiological, clinical, and paraclinical data was filled out for all patients:

Participants were recruited using non-probability consecutive sampling, which included all eligible diabetic patients who presented during the study period. This was therefore not a convenience sample.

A preliminary calculation of the minimum sample size was performed using Cochran’s formula, resulting in a minimum required sample size of 138 participants. However, in accordance with the adopted recruitment strategy, the consecutive inclusion of all eligible patients was continued throughout the predefined study period, resulting in a total of 267 participants, thus exceeding the minimum required sample size.

All patients underwent a general clinical examination, a foot examination, and a neurological examination (examination of sensitivity and motor function, search for osteotendinous reflexes and gait disorders…).

For the diagnosis of distal symmetric polyneuropathy (DSPN), the Michigan score or Michigan Neuropathy Screening Instrument (MNSI) [4] was used. The monofilament, reflex hammer, and tuning fork were used for the physical examination. The MNSI is widely used for the detection of DSPN in diabetes. It includes two distinct assessments:

MNSI with self-assessment by the patient: this is a 15-item self-administered questionnaire. A score ≥ 7 is considered pathological [8].Physical MNSI: this is a clinical assessment based on the examination of the lower limbs, looking for deformities, dry skin, calluses, infections, fissures, and ulcers. Ankle reflexes are also elicited. Vibration sensation is also tested in the big toe.

A score ≥ 2.5 is considered pathological [8].

In this study, we also relied on the DN4 score.

A score ≥ 4 confirmed the diagnosis of painful polyneuropathy.

Patients were considered to have distal symmetric polyneuropathy (DSPN) when:

The clinical MNSI score was ≥2.5 and/or the MNSI questionnaire score was ≥7 and/orThe DN4 score was ≥4 was incorporated into the overall diagnosis of DSPN

This dual approach increased the reliability of DPN screening in our study population.

The risk of risk lesions according to the International Working Group on the Diabetic Foot was used.

The analytical study is divided in this work into two parts: a univariate study of factors associated with DSPN and a multivariate study of independent predictive factors of DSPN.

Two groups of patients were determined: those who developed DSPN (G1: n = 125) or not (G2: n = 142).

The data collected were entered directly into an input form developed in Excel and analyzed using the Epi-info 7.2.2.6 software. Qualitative variables were expressed as absolute (n) or relative (%) frequencies. Quantitative variables were expressed as mean (m) ± standard deviation at 95%.

Variables with a p-value < 0.20, those with a p-value < 0.05 in the univariate analysis, and those of clinical interest according to the literature were entered into a multivariate logistic regression model to identify factors independently associated with DSPN.

The final model was obtained using backward stepwise regression, taking into account both the statistical significance and the clinical relevance of the variables.

For logistic regression, the reference categories corresponded to the modalities considered to be at lower risk or used for clinical comparison (e.g., absence of hypertension, shortest duration of diabetes, youngest age group, absence of neuropathy).

A preliminary check for multicollinearity between HbA1c and fasting/capillary blood glucose was performed. The indices were acceptable (covariance = 0.845; VIF = 1.183), allowing for the simultaneous inclusion of both variables in the logistic regression model.

The Odds Ratios (OR) calculated from univariate and multivariate logistic regression were presented with their 95% confidence intervals as well as their significance tests. For the comparison of percentages, the chi-squared test was used along with Fisher’s exact test. To check the significance of the OR, the parametric Wald test is used to test the role of explanatory variables. A difference was considered significant when the p-value was less than 5%.

This study was conducted in accordance with ethical principles.

Administrative approval was obtained from the administration of the Yopougon Attié General Hospital. Informed consent was obtained from each participant. Participation was voluntary, without financial compensation, and patients could withdraw at any time without any impact on their care. The anonymity and confidentiality of the data were guaranteed; the information collected was used exclusively for scientific purposes.

In our study, the prevalence of distal symmetric polyneuropathy (DSPN) was 46.82%, diagnosed using the DN4 and MNSI scores. This high prevalence confirms that DSPN is a common complication among diabetic patients followed in outpatient care in our hospital setting. Our prevalence of 46.82% is comparable to that reported by the major meta-analysis by Shiferaw et al. [3], which estimated the overall prevalence of diabetic peripheral neuropathy in Africa at 46%. In this same analysis, the prevalence in West Africa was 49.4%, a value higher than that observed in our study.

Several African studies, however, report prevalences higher than ours. In Côte d’Ivoire, Yao A et al. [15] found a prevalence of 56.5% in an elderly population. Similarly, Elleuch et al. [9], in Tunisia reported a prevalence of 57.8%, while Traoré S in Mali found a prevalence of 69.8% [10]. Our prevalence is also lower than that found in Benin by Gnonlonfoun [11], which was SS57%. These prevalences higher than ours could be explained by differences in the diagnostic tools used.

Conversely, our prevalence is higher than those found by Camara A, which is 28.8% [12], and that of Nientao I, which is 28.5% [13]. It is also higher than that reported by Djibril Y, which is 43.20% [14], highlighting a heterogeneity of results within the West African region itself.

Overall, the prevalence observed in our study reflects the reality of this microangiopathic complication among African diabetic patients, exposing them to a high risk of diabetic foot ulcers.

Compared to international data, the prevalence of DSPN in our study was higher. Indeed, Juan S et al., in a meta-analysis, reported a prevalence of 30%. A similar prevalence (30%) was found by Wang L et al., [16]. Moreover, the EURODIAB IDDM study, conducted in 16 European countries and including 3250 patients, reported a prevalence of 26% [17].

In our study, the average age of diabetic patients was 55.78 ± 12.40 years, with 69.66% of patients over 50 years old. Our result is close to that reported by Traoré S [10], which was 55 years, and that of Sangaré S, which was 55 years ± 13.9 [18]. It is also close to that found by Djibril Y, which was 52.63 ± 11.32 years [14]. However, it is lower than the average age reported by Elleuch et al., in Tunisia, which was 62.07 years [9], and by Barbosa et al., in Portugal, which was 69 ± 9.1 years [19], indicating a delay in PNDS screening in these countries.

1)Age>50yearsasapredictivefactorfordistalandsymmetricpolyneuropathy

In our study, among the sociodemographic factors evaluated (sex, economic status, average age), only age divided into groups was found to be associated with distal and symmetric polyneuropathy (DSPN) in bivariate analysis (p = 0.043). Age groups over 50 years had the highest frequency of DSPN, suggesting a growing relationship between age and the occurrence of neuropathy. Our results corroborate those of Oulad Sayad et al. [20], Djorolo et al. [21], as well as Jember [22].

After adjusting for other factors in the multivariate logistic regression, age over 50 years appeared as an independent predictive factor for DSPN. Patients aged 50 to 59 years had the highest risk (adjusted OR = 6.45; 95% CI [2.13 - 19.51]) compared to the reference group (≤40 years), while the risk remained significant in patients ≥ 60 years (adjusted OR = 2.29; 95% CI [1.30 - 18.74]), suggesting a risk peak in the sixties. These results confirm the well-established trend of increased prevalence of diabetic neuropathy after age 50.

Comparison with international literature, however, shows threshold variations. Our threshold of 50 years corresponds to that found by Jember G et al. in Ethiopia [22], while other studies in China conducted by Cheng Y et al. [23], and in Uganda by Kisozi T et al. [24], identified higher thresholds of 66 and 60 years, respectively. These differences can be explained by the specific characteristics of the populations studied, such as life expectancy, access to healthcare, methods of NCD diagnosis, or the presence of distinct genetic and environmental factors.

From a pathophysiological standpoint, the association between advanced age and DPN results from several synergistic mechanisms: progressive decline in nerve regeneration capacity, accumulation of oxidative stress, and increased susceptibility to microvascular damage, which, combined with the deleterious effects of chronic hyperglycemia, make the peripheral nervous system more vulnerable with age.

2)Diabeticretinopathy,apredictivefactorofdistalandsymmetricalpolyneuropathy.

In our study, diabetic retinopathy was significantly associated with distal and symmetrical polyneuropathy (DSPN) in bivariate analysis (p = 0.003). This association is well documented in the literature, both at the African and international levels, and reflects the tendency of diabetic microvascular complications to coexist in the same patient due to common pathophysiological mechanisms. Studies conducted in Morocco by Kacem et al. [25] and Berkia et al. [26] showed a significant association between diabetic retinopathy and peripheral neuropathy. The same result was observed in the European EURODIAB study by Tesfaye et al. [17].

After adjusting for other variables in the multivariate logistic regression, the presence of diabetic retinopathy remains an independent predictive factor for PNDS (adjusted OR = 3.08; 95% CI [1.40 - 6.76]; p = 0.005). Our results corroborate those of Elleuch M et al. in Tunisia, who also found diabetic retinopathy to be a predictive factor for DSPN [9].

3)Diabetesduration>10yearspredictiveofdistalandsymmetricalpolyneuropathy

In our study, the association between the duration of diabetes and distal and symmetrical polyneuropathy (DSPN) was highlighted in a bivariate analysis, with patients having diabetes for more than 10 years being the most affected (p = 0.040). Our results corroborate numerous studies, notably those by Malgrange D [27], and Aouiche et al. [28], which reported a positive correlation between the duration of diabetes and the occurrence of distal neuropathy.

In multivariate analysis, a duration of diabetes between 10 and 19 years appeared as an independent predictive factor for DSPN (adjusted OR = 2.95; 95% CI [1.10 - 7.93]; p = 0.031), while the other classes lost their significance after adjustment. This result confirms the consistency of diabetes duration as a predictive factor, reported both in the African and international literature. Our result is comparable to that of Gebabo T et al. [29], who identified a diabetes duration of more than 10 years as a predictor of DSPN. In Ethiopia, Mekuria Negussie Y. et al. [30] identified a diabetes duration > 5 years as an independent factor. In China, Li L etal. [31], as well as a meta-analysis conducted by Liu X et al. [32], including 12,116 patients, also confirmed the duration of diabetes as a major predictor of peripheral neuropathy.

4)TherapeuticadherenceasapredictivefactorofDSPN

In our study, poor or moderate therapeutic adherence, evaluated using the Morisky score, was significantly associated with distal symmetric polyneuropathy (DSPN) in bivariate analysis (p = 0.001). Our results are consistent with those reported in the literature, where several studies emphasize the major impact of treatment adherence on the occurrence of diabetic complications. For example, Badr W et al. [33] showed that poor adherence was significantly associated with peripheral neuropathy, and Khunti K et al. [34] reported an increased risk of micro- and macrovascular complications in non-adherent patients.

After adjusting for other variables in the multivariate logistic regression, therapeutic adherence appeared as an independent predictive factor for DSPN. Compared to adherent patients, those with moderate adherence (aOR = 2.37; p = 0.021) or poor adherence (aOR = 2.54; p = 0.041) had a significantly higher risk of developing DSPN. Our results are consistent with international studies: in Ethiopia, Demoz et al. [35], using the Morisky score in an adjusted logistic model, showed that poor adherence significantly increased the risk of peripheral neuropathy after adjustment. In India, Samu A et al. [36] observed that good adherence reduced the severity of sensory neuropathy.

From a pathophysiological standpoint, this association could be explained by the central role of therapeutic adherence in glycemic control. Poor adherence to treatment promotes chronic hyperglycemia involved in the development of diabetic neuropathy.

5)GlycemicimbalanceasapredictivefactorofDSPN

In our study, elevated fasting blood glucose as well as glycated hemoglobin (HbA1c) above 7% were significantly associated with the occurrence of distal symmetric polyneuropathy (DSPN) in univariate analysis. This result is consistent with the data in the literature. Indeed, Alshammari et al. [37] identified fasting blood glucose as being associated with DSPN, while Hartmann et al. [38] reported a significantly higher prevalence of diabetic polyneuropathy in patients with high HbA1c levels, suggesting a close link between poor chronic glycemic control and this complication. After adjustment for potential confounding factors using multivariate logistic regression, high fasting blood glucose and an HbA1c above 7% remained independent predictive factors of DSPN.

This result highlights that glycemic imbalance, whether assessed by immediate or long-term integrated markers, has an impact on the occurrence of diabetic neuropathy. Our results corroborate the work of Pai YW et al. [39], and Wu et al. [40], who demonstrated the independent predictive role of glycemic parameters in DSPN. Intensive glycemic treatment is necessary to reduce the risk of DSPN by 64%, as reported by the DCCT study (Diabetes Control and Complications Trial) [41].

Limitations of the Study

This study has certain limitations, including its single-center, outpatient setting and its cross-sectional design, which limit the generalizability of the results and preclude any causal inferences. The exclusion of non-ambulatory patients may have introduced selection bias. Finally, the lack of confirmatory nerve conduction studies constitutes a diagnostic limitation.