Serum nerve growth factor and lipid-derived indices in Iraqi adults with diabetic peripheral neuropathy: a hospital-based case-control study at the specialized center for endocrinology and diabetes, Al-Rusafa, Baghdad

¹Department of Chemistry, College of Science, University of Baghdad, Baghdad, Iraq

^&Corresponding author
Shymaa Mohamad, Department of Chemistry, College of Science, University of Baghdad, Baghdad, Iraq

Introduction    

Diabetic peripheral neuropathy (DPN) is one of the most clinically important complications of type 2 diabetes mellitus (T2DM). It is associated with sensory loss, neuropathic pain, gait disturbance, foot ulceration, risk of lower-limb amputation, and reduced quality of life [1]. Although persistent hyperglycemia remains central to its development, DPN is now understood as a multifactorial disorder involving oxidative stress, microvascular dysfunction, inflammation, mitochondrial injury, dyslipidemia, and impaired neuronal repair [2,3]. These mechanisms may progress silently before advanced neurological disability becomes clinically apparent; therefore, biomarkers that reflect neuronal injury and metabolic disturbance may improve understanding of DPN in specific populations [4]. Nerve growth factor (NGF) is a neurotrophin required for the survival, maintenance, and regeneration of peripheral sensory and sympathetic neurons [5]. Experimental and clinical evidence suggests that reduced neurotrophic support may contribute to axonal degeneration, Schwann-cell dysfunction, and impaired nerve regeneration in diabetes [6,7]. Serum NGF has therefore attracted attention as a possible marker of diabetic nerve injury. However, reported findings differ across populations, disease stages, and diagnostic approaches, and the behavior of NGF in relation to obesity among patients who already have DPN is still not well defined [8].

Lipid disturbance is another pathway that may contribute to diabetic nerve damage [9,10]. Elevated triglycerides, increased low-density lipoprotein cholesterol, reduced high-density lipoprotein cholesterol, and atherogenic dyslipidemia may aggravate endothelial dysfunction, oxidative stress, and impaired microvascular blood supply to peripheral nerves [11,12]. Lipid-derived indices, including the Atherogenic index of plasma (AIP), Triglyceride-glucose (TyG) index, Castelli risk index-I (CRI-I), Castelli risk index-II (CRI-II), and cholesterol index, combine lipid and glycemic information and may reflect cardiometabolic burden better than isolated lipid values [13-16]. Nevertheless, NGF and lipid-derived indices are often evaluated separately, leaving uncertainty about their combined pattern among adults with DPN. In Iraq, evidence on serum NGF and lipid-derived indices among adults with electrophysiologically confirmed DPN is limited. This gap is important because local patient characteristics, glycemic control, obesity patterns and lipid abnormalities may differ from those reported elsewhere. The present hospital-based case-control study was conducted at the Specialized Center for Endocrinology and Diabetes in Al-Rusafa, Baghdad. The aim was to compare serum NGF, conventional lipid parameters and lipid-derived indices between adults with DPN and controls without diabetes or neuropathic symptoms. A secondary objective was to examine whether obese DPN cases differed from non-obese DPN cases in these biomarkers. The study addressed three main research questions: whether adults with DPN have lower serum NGF than controls; whether adults with DPN have higher conventional lipid parameters and lipid-derived indices than controls; and whether obesity among DPN cases is associated with additional differences in NGF, lipid profile or lipid-derived indices.

Methods     

Study design and setting: this was a hospital-based case-control study conducted at the Specialized Center for Endocrinology and Diabetes in Al-Rusafa, Baghdad, Iraq, between January 2024 and December 2024. The study was prospectively set up for research purposes. Participants were invited, consented, clinically assessed, and sampled specifically for this study; the analysis was not a secondary analysis of routine medical records. Cases were adults with T2DM and electrophysiologically confirmed DPN. Controls were adults without diabetes and without neuropathic symptoms recruited during the same period. The same fasting blood-sampling procedure, biochemical testing method, serum NGF assay, and lipid-index calculations were applied to cases and controls.

Study population, sampling, and sample-size rationale: the study included 62 adults aged 40-70 years: 42 cases with T2DM and electrophysiologically confirmed DPN and 20 controls without diabetes or neuropathic symptoms. The age range was selected because DPN is clinically relevant in middle-aged and older adults with T2DM, whereas younger adults are less likely to have established DPN, and adults above 70 years are more likely to have age-related neuropathy, frailty, vascular disease, or other non-diabetic causes of peripheral nerve dysfunction. Cases were recruited consecutively: every patient attending the center during the study period who met the inclusion criteria, had electrophysiological confirmation of DPN, had no exclusion criterion, completed fasting blood sampling, and consented to participate was invited. Controls were recruited during the same period using the same consecutive approach among adults who met the control definition. The sample size was feasibility-based because the study required electrophysiological confirmation, fasting biochemical testing, and serum NGF enzyme-linked immunosorbent assay measurement within one center and one study period. The study therefore included all eligible and consenting participants who completed the full protocol during the recruitment window.

Case definition, control definition and BMI stratification rationale: cases were defined as adults with T2DM and DPN confirmed by electromyography and nerve conduction studies [17]. Controls were adults without diabetes and without neuropathic symptoms. The primary case-control comparison addressed whether DPN cases differed from controls in serum NGF and lipid-derived indices. The BMI stratification of cases into D1 (BMI <30 kg/m²) and D2 (BMI ≥30 kg/m²) was a prespecified secondary within-case analysis. It was included because obesity is biologically linked to insulin resistance, dyslipidemia, chronic low-grade inflammation, and microvascular dysfunction, all of which may influence DPN-related metabolic and neurotrophic markers [8,18]. This stratification addressed the research question of whether obese DPN cases differed from non-obese DPN cases in NGF, lipid profile, and lipid-derived indices. It did not replace the main case-control design.

Eligibility criteria and data collection: participants were included if they were aged 40-70 years and met the definition for either case or control status. Cases had physician-confirmed T2DM and DPN confirmed by electrophysiological testing. Controls had no diabetes and no neuropathic symptoms. Exclusion criteria were type 1 diabetes, chronic liver or kidney disease, pregnancy, current insulin therapy, conditions likely to affect neuropathy or lipid metabolism, and failure to complete an 8-hour fast before blood sampling. Weight and height were measured using calibrated equipment, and BMI was calculated as kg/m². Venous blood was collected after an 8-10 hour fast. Fasting blood sugar, glycated hemoglobin, triglycerides, total cholesterol, and high-density lipoprotein cholesterol were measured using Roche diagnostic kits on a Cobas C111 automated analyzer. Low-density lipoprotein cholesterol and very-low-density lipoprotein cholesterol were calculated using the Friedewald equation when triglycerides were below 400 mg/dL [19]. Serum NGF was measured using a sandwich enzyme-linked immunosorbent assay according to the manufacturer´s instructions.

Definitions and derived indices: diabetic peripheral neuropathy was defined as clinically suspected peripheral neuropathy in a patient with T2DM confirmed by electromyography and nerve conduction studies [17]. Body mass index stratification was used only for the planned within-case comparison. Atherogenic index of plasma was calculated as log10 (triglycerides/high-density lipoprotein cholesterol) after conversion to mmol/L [13,16]. The TyG Index, CRI-I, CRI-II, and Cholesterol Index were calculated from fasting glucose and lipid parameters according to their standard formulas [14-16]. All indices were interpreted as research variables describing metabolic risk patterns, not as diagnostic thresholds.

Bias and comparability: the same blood-collection procedures, biochemical methods, NGF assay protocol, and lipid-index formulas were applied to cases and controls to reduce measurement bias. Electrophysiological confirmation was used to reduce misclassification of DPN among cases. Controls were screened for absence of diabetes and neuropathic symptoms. Selection bias remains possible because recruitment was hospital-based and non-random. Residual confounding is also possible because the study did not include individual matching or multivariable adjustment.

Statistical analysis: data were analyzed using IBM SPSS Statistics version 22. Continuous variables were summarized as mean ± standard deviation. The primary case-control analysis compared controls with DPN cases, presented as non-obese and obese case subgroups because BMI stratification was prespecified. Serum NGF, glycemic variables, conventional lipid parameters, and lipid-derived indices were compared across controls, non-obese DPN cases, and obese DPN cases using one-way analysis of variance. This approach allowed assessment of the main case-control contrast and the planned secondary comparison between obese and non-obese DPN cases. Pearson correlation coefficients were calculated separately within non-obese and obese DPN cases to explore whether NGF and lipid-derived indices were associated with glycemic and lipid variables among cases. A two-sided P-value below .05 was considered statistically significant. No multivariable logistic regression was performed because the sample size and number of controls were too small for stable adjusted case-control modeling. Findings were interpreted as exploratory associations consistent with STROBE reporting principles for case-control studies [20,21].

Ethical consideration: ethical approval was obtained from the Research Ethics Committee of the College of Science, University of Baghdad (Approval No. CSEC/0325/0044; 13 March 2025). Written informed consent was obtained from all participants before enrollment. Participant confidentiality was maintained, and the study was conducted in accordance with the Declaration of Helsinki.

Results     

Participants included in the case-control analysis and baseline comparability: all 62 enrolled participants completed the required assessments and were included in the final analysis: 42 cases with T2DM and electrophysiologically confirmed DPN and 20 controls without diabetes or neuropathic symptoms. The case-control component remained the main comparison throughout the analysis. Cases were further stratified into 21 non-obese DPN cases (D1) and 21 obese DPN cases (D2) for the prespecified secondary within-case obesity comparison. Age did not differ significantly among controls (54.10 ± 6.20 years), D1 cases (56.30 ± 7.10 years), and D2 cases (55.80 ± 6.80 years; P=.542), supporting broad age comparability. Body mass index differed significantly across groups, as expected from the planned BMI-based case stratification (24.30 ± 2.10, 28.10 ± 1.20, and 34.80 ± 3.10 kg/m² respectively; P<.001). Glycemic status differed by case-control status: FBS and HbA1c were higher in both DPN case groups than in controls (P<.001). These findings are presented in Table 1, Table 2.

Serum NGF among cases and controls: serum NGF differed significantly across the study groups (P<.001). Mean NGF was lower in non-obese DPN cases (82.30 ± 15.70 pg/mL) and obese DPN cases (79.80 ± 17.10 pg/mL) than in controls (145.60 ± 18.40 pg/mL). The two DPN subgroups did not differ significantly from each other. Thus, the main difference in NGF was observed between DPN cases and controls, rather than between BMI-defined DPN case subgroups. These findings are shown in Table 3.

Lipid profile and lipid-derived indices among cases and controls: diabetic peripheral neuropathy cases had a more adverse lipid pattern than controls. Total cholesterol was higher in non-obese DPN cases (221.60 ± 34.50 mg/dL) and obese DPN cases (228.30 ± 38.20 mg/dL) than in controls (165.40 ± 22.10 mg/dL; P<.001). Triglycerides were also higher in non-obese DPN cases (198.70 ±52.10 mg/dL) and obese DPN cases (212.40 ± 58.60 mg/dL) than in controls (108.50 ± 24.30 mg/dL; P<.001). High-density lipoprotein cholesterol (HDL-C) was lower in both DPN groups than in controls, while LDL-C and VLDL-C were higher in both DPN groups than in controls. Lipid-derived indices followed the same pattern. Atherogenic index of plasma, TyG Index, CRI-I, CRI-II and Cholesterol Index were significantly higher in both DPN case groups than in controls (P<.001). These findings are summarized in Table 3,Table 4.

Associations between NGF and glycemic markers among diabetic peripheral neuropathy cases: within non-obese DPN cases, serum NGF was not significantly associated with BMI, fasting blood sugar or HbA1c. Within obese DPN cases, serum NGF was negatively associated with fasting blood sugar (r=-0.448; P=.041) and HbA1c (r=-0.541; P=.011), while its association with BMI was not statistically significant. These exploratory findings suggest that poorer glycemic control was associated with lower NGF only among obese DPN cases in this sample. The results are presented in Table 5.

Associations between lipid-derived indices and metabolic variables among DPN cases: among non-obese DPN cases, AIP was positively associated with TyG index, fasting blood sugar, triglycerides, CRI-I and CRI-II. TyG Index was also positively associated with AIP, fasting blood sugar, triglycerides, CRI-I and CRI-II. Among obese DPN cases, TyG Index was positively associated with fasting blood sugar and triglycerides, while AIP did not show statistically significant associations with the selected metabolic variables. These results indicate stronger clustering of lipid-derived indices with glycemic and lipid variables in non-obese DPN cases than in obese DPN cases, although these subgroup findings should be interpreted cautiously because of the small sample size. The findings are presented in Table 6,Table 7.

Discussion     

This hospital-based case-control study found that adults with T2DM and electrophysiologically confirmed DPN had lower serum NGF and a more adverse glycemic and lipid profile than controls without diabetes or neuropathic symptoms. Both obese and non-obese DPN cases had lower NGF and higher FBS, HbA1c, total cholesterol, triglycerides, LDL-C, VLDL-C, AIP, TyG Index, CRI-I, CRI-II, and Cholesterol Index than controls, while HDL-C was lower. The secondary within-case analysis showed no statistically significant difference in NGF or lipid-derived indices between obese and non-obese DPN cases. The main pattern in the data was therefore a case-control difference between adults with DPN and controls, rather than an obesity-related difference among DPN cases. The lower NGF concentrations observed among DPN cases are consistent with the biological role of NGF in peripheral nerve survival and repair [5]. Hyperglycemia, oxidative stress, and microvascular dysfunction may reduce neurotrophic support, impair Schwann-cell function, and limit axonal regeneration [6,22]. Previous work has suggested that patients with diabetic neuropathy may have reduced neurotrophic factor availability, although reported values vary by population, assay method, neuropathy definition, and disease severity [7]. In the present study, reduced NGF was seen in both DPN subgroups when compared with controls, supporting the view that impaired neurotrophic signaling may be linked to established DPN. However, because the study was case-control and cross-sectional at enrollment, it cannot determine whether lower NGF preceded neuropathy or resulted from established nerve damage.

The lipid findings also support a metabolic pattern compatible with diabetic nerve injury. Diabetic peripheral neuropathy cases had higher triglycerides, total cholesterol, LDL-C, VLDL-C, and lipid-derived indices, together with lower HDL-C. Dyslipidemia may contribute to peripheral nerve damage through endothelial dysfunction, inflammatory activation, oxidative injury, and reduced microvascular perfusion [9,12, 23]. The high AIP and TyG Index values among cases suggest that atherogenic lipid burden and insulin-resistance-related metabolic stress were more pronounced in participants with DPN than in controls [13-16]. These indices should not be interpreted as independent diagnostic tests in this study, because they are calculated from the same lipid and glucose measurements. However, their consistent elevation among cases strengthens the observation that DPN in this sample was accompanied by combined glycemic and lipid abnormalities. The BMI-stratified findings require careful interpretation. Obesity is biologically linked to insulin resistance, dyslipidemia, and inflammation, and may contribute to neuropathy risk [8,18]. For this reason, the study included a planned secondary comparison of obese and non-obese DPN cases. Nevertheless, obese DPN cases did not differ significantly from non-obese DPN cases in NGF or lipid-derived indices. This may reflect the small sample size, the presence of established neuropathy in both case subgroups, or the dominant influence of diabetes and glycemic control over BMI category in this study. The findings should not be interpreted as evidence that obesity is unrelated to DPN; rather, they show that this study did not detect an additional obesity-related biomarker difference among adults who already had DPN. This study has several strengths. It was prospectively organized for research purposes, used a clear hospital-based case-control design, recruited cases and controls during the same period, confirmed DPN electrophysiologically among cases, applied standardized fasting biochemical procedures, and evaluated NGF together with conventional and lipid-derived metabolic markers. The study also has important limitations. It was conducted in a single center, used a small feasibility-based sample, applied consecutive rather than random sampling, did not include a diabetic control group without neuropathy, and did not perform individual matching or multivariable adjustment. Potential confounders such as diabetes duration, medication use, renal function, inflammatory status, vitamin status, smoking, diet, and physical activity were not fully controlled. The findings should therefore be interpreted as exploratory associations. Larger multicenter case-control or cohort studies, including diabetic patients without neuropathy, are needed to confirm whether NGF and lipid-derived indices are independently associated with DPN presence, severity, or progression.

Conclusion     

This hospital-based case-control study showed that Iraqi adults aged 40-70 years with electrophysiologically confirmed DPN had lower serum NGF and higher glycemic, lipid and lipid-derived metabolic indices than controls without diabetes or neuropathic symptoms. The secondary BMI-stratified analysis did not show significant differences in NGF or lipid-derived indices between obese and non-obese DPN cases. These findings suggest that DPN in this study population was associated with combined neurotrophic and metabolic abnormalities, mainly reflected by the case-control contrast. Because of the single-center setting, feasibility-based sample and absence of adjusted modeling, the findings should be interpreted cautiously and confirmed in larger studies.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Conception and study design: Shaymaa Mohammed Al- Sumaydae, Nada Abdul Kareem Kadhim. Data collection: Shaymaa Mohammed Al- Sumaydae. Laboratory investigation: Shaymaa Mohammed Al- Sumaydae. Data analysis and interpretation: Shaymaa Mohammed Al- Sumaydae, Nada Abdul Kareem Kadhim. Manuscript drafting: Shaymaa Mohammed Al- Sumaydae. Manuscript revision: Shaymaa Mohammed Al- Sumaydae, Nada Abdul Kareem Kadhim. Guarantor of the study: Nada Abdul Kareem Kadhim. All authors approved the final version of the manuscript.

Acknowledgments     

The authors gratefully acknowledge the staff of the Specialized Center for Endocrinology and Diabetes in Al-Rusafa, Baghdad, for their assistance with participant recruitment and clinical assessments. Appreciation is also extended to the laboratory personnel for their support in biochemical and immunological analyses.

Tables      

Table 1: flow of cases and controls through the hospital-based case-control study

Table 2: baseline characteristics of controls and diabetic peripheral neuropathycases stratified by body mass index

Table 3: case-control comparison of serum nerve growth factor and conventional lipid parameters

Table 4: case-control comparison of lipid-derived indices in controls and body mass index-stratified diabetic peripheral neuropathy cases

Table 5: within-case associations between serumnerve growth factor and body mass index or glycemic markers among body mass index-stratified diabetic peripheral neuropathy cases

Table 6: within-case associations between atherogenic index of plasma and metabolic variables among body mass index-stratified diabetic peripheral neuropathycases

Table 7: within-case associations between TyG index and metabolic variables among body mass index-stratified diabetic peripheral neuropathy cases

References