Inflammatory bowel disease (IBD) has systemic consequences that extend beyond the gastrointestinal tract. While nailfold capillaroscopy (NFC) is widely utilized in many rheumatological disorders, its significance in evaluating microvascular changes in IBD remains unknown. This study aimed to standardize NFC use in IBD patients and relate NFC results to clinical and laboratory criteria.
MethodsThis observational case–control study included histology- and colonoscopy-diagnosed IBD patients. We performed NFC on eight fingers of each subject using a Dino-Lite digital microscope, adhering to European Alliance of Associations for Rheumatology guidelines. Capillary characteristics, including density, morphology, and dimensions, as well as microhemorrhages, were assessed at 200×.
ResultsThe study included 50 IBD patients and 30 healthy controls. IBD patients had significantly lower capillary density (7.52±0.68capillaries/mm) than controls (8.18±0.63capillaries/mm, p<0.001), with 26% of IBD patients exhibiting densities below seven capillaries/mm. In IBD patients, microhemorrhages were more prevalent than in controls (p=0.032). Raynaud's phenomenon was more commonly detected in patients with low mean capillary density (p=0.04).
ConclusionIBD patients had reduced mean capillary density and increased microhemorrhages. These changes suggested that NFC could be a non-invasive way provides unique insights into IBD microvascular health.
.
La enfermedad inflamatoria intestinal (EII) tiene consecuencias sistémicas que se extienden más allá del tracto gastrointestinal. Si bien la capilaroscopia del pliegue ungueal (CPU) se utiliza ampliamente en muchos trastornos reumatológicos, se desconoce su importancia para evaluar los cambios microvasculares en la EII. Este estudio tuvo como objetivo estandarizar el uso de la CPU en pacientes con EII y relacionar los resultados con criterios clínicos y de laboratorio.
MétodosEste estudio observacional de casos y controles incluyó pacientes con EII diagnosticados mediante histología y colonoscopia. Se realizó la CPU en ocho dedos de los sujetos utilizando un microscopio digital Dino-Lite, siguiendo las directrices de la Alianza Europea de Asociaciones de Reumatología. Las características capilares, incluida la densidad, la morfología y las dimensiones, así como las microhemorragias, se evaluaron a 200×.
ResultadosEl estudio incluyó a 50 pacientes con EII y 30 controles sanos. Los pacientes con EII presentaron una densidad capilar significativamente menor (7,52±0,68 capilares/mm) que los controles (8,18±0,63 capilares/mm, p<0,001), con un 26% de los pacientes con EII presentando densidades inferiores a siete capilares/mm. En los pacientes con EII, las microhemorragias fueron más frecuentes que en los controles (p=0,032). El fenómeno de Raynaud se detectó con mayor frecuencia en pacientes con baja densidad capilar media (p=0,04).
ConclusiónLos pacientes con EII presentaron una densidad capilar media reducida y un aumento de las microhemorragias. Estos cambios sugirieron que la CPU podría ser una forma no invasiva de brindar información única sobre la salud microvascular de la EII.
Inflammatory bowel disease (IBD), encompassing Crohn's disease (CD) and ulcerative colitis (UC), is a chronic inflammatory condition primarily involving the gastrointestinal tract but increasingly recognized to have systemic implications.1 IBD is associated with a wide range of extraintestinal manifestations (EIMs), including musculoskeletal, dermatologic, and vascular complications. These symptoms are believed to result from chronic inflammation that affects organs and systems outside the gut, therefore aggravating symptoms in individuals with IBD and posing difficulties for their complex illness management.2,3
Microvascular abnormalities are increasingly manifesting in autoimmune and chronic inflammatory illnesses.4 Vascular endothelial cells and microvascular alterations are recognized as crucial elements in IBD pathophysiology.5 Nailfold capillaroscopy (NFC) is a non-invasive method commonly applied in rheumatology that shows early microvascular changes in diseases like systemic sclerosis (SSc), rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and seronegative spondyloarthropathy. Visualizing capillary shape, density, and structure at the nailfold has helped to identify microvascular impairment before clinical symptoms start.6–9 Nevertheless, NFC results in IBD populations remain understudied, despite the documented systemic character of IBD and reports of vascular impairment in these individuals.
Evolving data indicate that, like other inflammatory diseases, NFC may expose notable capillary abnormalities in IBD patients, including reduced capillary density and capillary morphologic changes.10 These capillaroscopic abnormalities could potentially serve as markers of systemic disease impact in IBD, correlating with disease activity, inflammatory markers, and clinical symptoms.11,12 However, the limited number of existing studies necessitates the development of standardized methodologies to evaluate NFC parameters in IBD, thereby enhancing our understanding of the extent and clinical significance of these vascular changes. The study protocol aimed to assess microvascular abnormalities in IBD patients using NFC, and to explore associations between capillaroscopic findings and clinical as well as laboratory parameters.
Patients and methodsStudy design and patient recruitmentThis observational case–control study was conducted at a university hospital. Patients were recruited from the IBD clinic in the Department of Tropical Medicine and Gastroenterology then referred to the NFC unit in the Department of Rheumatology, Rehabilitation and Physical Medicine, aiming to evaluate NFC results and explore their associations with clinical and laboratory parameters as well as disease activity indices. The study was conducted from February 2024 to December 2024. The study included 50 IBD patients and 30 age and gender frequency matched healthy participants. Each patient was matched with 0.6 controls. Healthy controls were recruited from hospital staff and patients’ companions, excluding relatives of patients to avoid genetic influence.
Sample size was calculated by the open EPI software based on the difference between CD patients and healthy controls regarding capillary density in a previous study.10
The study received approval from the Institutional Review Board of the Faculty of Medicine, Assiut University, Egypt, in accordance with the Declaration of Helsinki (IRB NO.: 04-2024-300358). All participants signed a written informed consent following careful discussion of study's goal, methods, and confidentiality policies.
Eligibility criteria: Adults aged 18–70 with a confirmed IBD (UC or CD) diagnosis based on endoscopy and histopathology.
Exclusion criteria (applied to both patients and controls): Comprise the presence of other autoimmune or connective tissue diseases, such as SLE or SSc, comorbidities affecting microvascular health including (diabetes mellitus, cardiovascular disease, chronic renal disease, or recent infections), nailfold trauma, smoking, pregnancy, hormonal medications or vasoactive drugs intake and inability to provide an informed consent.
Sociodemographic, clinical and laboratory assessmentInitial data collection at the IBD clinic included demographics, disease characteristics, laboratory investigations and medications. Laboratory data included inflammatory markers (erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP)), serum albumin, and complete blood count (CBC). Disease activity in UC was measured using the Mayo score, composed of four items (bleeding, stool frequency, physician assessment, and endoscopic appearance) rated from 0 to 3. The four individual scores are summed to yield a final score that ranges from 0 to12, with higher scores indicating more severe disease.13 Remission is defined as a total score of ≤2 with no individual score >1, and a rectal bleeding score of 0.14 CD activity was measured by the Crohn's Disease Activity Index (CDAI) which consists of eight items (number of liquid or soft stools each day, abdominal pain, general wellbeing, presence of complication, taking antidiarrheal drugs such as diphenoxylate or loperamide, presence of an abdominal mass, hematocrit of <0.47 in men and <0.42 in women and percentage deviation from standard weight). The CDAI categorizes patients into asymptomatic remission (CDAI <150), mild-to-moderate (150–220), moderate-to-severe (220–450), and severe-fulminant (>450).15
Rheumatological evaluation included obtaining the history of musculoskeletal symptoms, Raynaud's phenomenon (RP) and EIMs. Examination of the peripheral joints, back and sacroiliac joints was done. Identifying patients with RP was based on a three step clinical approach, relying on cold sensitivity, recurrent biphasic or triphasic color changes of the fingers triggered by cold exposure and association of these episodes with pain or paresthesia.16 In addition, skin was inspected to detect necrosis or scarring.
NFC techniqueA standardized NFC technique adhering to the European Alliance of Associations for Rheumatology (EULAR) study group on Microcirculation in Rheumatic Diseases6,7,17 was applied using Dino-Lite MEDL4N pro, video capillaroscope. All NFC examinations were performed by the same rheumatologist. Before examination, study participants were acclimatized for 20min at room temperature (20–24°C). At first, a drop of immersion oil was applied to the nailfold to improve resolution. The nailfold of all fingers was examined except for the thumbs. Six successive images were captured for each nailfold: two at 50× magnification to obtain a general view of the nailfold and four at 200× to assess the following parameters:
- •
Capillary density is calculated by counting capillary loops per linear mm in the distal row. A capillary density ≥7 capillaries per linear mm is identified as normal.7,11
- •
Microhemorrhages, interpreted as dark masses attributable to hemosiderin deposits within the distal row.7,17
- •
Capillary morphology was scored per capillary per image as normal (hairpin), nonspecific variation (tortuous or once or twice crossing) or abnormal (bushy, ramified, or bizarre-shaped).7,11
- •
Capillaries were classified according to their dimensions as follows: normal capillaries, with apical, afferent, and efferent limb diameters less than 20μm, dilated capillaries, with the diameter of all three limbs between 20 and 50μm, and giant capillaries, with the diameter of all three limbs greater than 50μm.6,18
We estimated the percentage of each capillary abnormality by dividing the total number of capillaries of that abnormality by the total number of capillaries in the distal row per image. Analysis of images was done via DinoCapture 2.0 software version 1.5.48.
The IBD group was further subdivided based on capillary density (normal ≥7capillaries/mm or low <7capillaries/mm) to assess the impact of different factors on capillary density.
Statistical analysisOpen Epi V.3.01 software was used to calculate the sample size. SPSS (version 26) was applied for data analysis. Categorical data were expressed as frequencies and percentages while numerical data were presented as mean±standard deviation/median (range). Student's t-test and Mann–Whitney U test were applied to compare means/medians among groups. The proportions were compared via the Chi-square and Fisher's exact test. The correlation was identified by using Spearman's or Pearson's correlation. Regression analysis was conducted to identify predictors of low capillary density and microhemorrhages including disease activity as a unified binary variable (active vs. remission). The p-value of less than 0.05 was considered significant. All tests were two-tailed.
ResultsDemographic and anthropometric data of IBD patientsThe IBD group comprised 34 females and 16 males, whereas the control group included 20 females and 10 males with a p-value=0.9. The mean age of IBD patients was 33.9±10.11 years, compared with 35.5±9.81 years in the control group (p=0.488). Similarly, weight, height and body mass index showed no significant differences between groups, with p-values of 0.056, 0.34 and 0.251, respectively.
Clinical and laboratory characteristics of IBD patientsThe clinical characteristics of IBD patients revealed that the majority of included patients had UC (90%), whereas a smaller proportion had CD (10%). The median disease duration was 5 years, ranging from 0.33 to 18 years. EIMs were present in 12% of patients, and 18% experienced RP. The median ESR was 21mm/h, and the median CRP was 11mg/dl. These findings underline the chronic inflammatory state in IBD patients and the presence of associated symptoms beyond intestinal inflammation, as shown in Table 1. Current treatment of IBD patients is illustrated in Table 2.
Clinical and laboratory characteristics of IBD patients.
| Variable | Mean±SD/median (range)/n (%) |
|---|---|
| Disease duration (years) | 5 (0.33–18) |
| Diagnosis | |
| • UC | 45 (90%) |
| Disease activity score median (range) | 4 (1–10) |
| • CD | 5 (10%) |
| Disease activity index | 292.4±93.98 |
| Extraintestinal manifestations | 6 (12%) |
| Raynaud's phenomenon | 9 (18%) |
| Laboratory investigations | |
| • Hemoglobin (g/dl) | 10.75±0.9 |
| • First hour ESR (mm/h) median (range) | 21 (5–60) |
| • CRP (mg/dl) median (range) | 11 (1–51) |
| • Serum albumin (g/dl) | 3.34±0.55 |
Data expressed as mean±SD/median (range) or frequency (%). UC: ulcerative colitis, CD: Crohn's disease, ESR: erythrocyte sedimentation rate, CRP: C-reactive protein.
As illustrated in Table 3, IBD patients had a lower mean capillary density (7.52±0.68capillaries/mm) compared to controls (8.18±0.63capillaries/mm), with a p-value <0.001. Additionally, 26% of IBD patients exhibited capillary densities below 7/mm, while none of the controls fell below this threshold (p=0.001). Microhemorrhages were more frequent in IBD patients (26%) than in controls (6.67%), with a statistically significant difference (p=0.032). Fig. 1 shows normal capillary density, distribution and shape in a healthy participant, whereas Fig. 2 demonstrates capillary abnormalities; microhemorrhage in a male patient with CD and a ramified capillary in a female patient with UC.
Nailfold capillaroscopic findings in IBD patients and the control group.
| Nailfold capillaroscopic parameters | IBD, n=50 | Controls, n=30 | p-Value |
|---|---|---|---|
| Mean capillary density/mm | 7.52±0.68 | 8.18±0.63 | <0.001 |
| Mean capillary density <7/mm, n (%) | 13 (26%) | 0 (0%) | 0.001 |
| Microhemorrhages, n (%) | 13 (26%) | 2 (6.67%) | 0.032 |
| Abnormal shape | |||
| • Bushy capillaries % median (range) | 0 (0–9.4) | 0 | 0.27 |
| • Bizarre-shaped % median (range) | 0 (0–1.88) | 0 (0–3.13) | 0.854 |
| • Ramified capillaries % median (range) | 0 (0–3.75) | 0 (0–5) | 0.738 |
| Abnormal dimensions | |||
| • Dilated capillaries % | 0 (0–12.5) | 0 (0–3.57) | 0.683 |
| • Giant capillaries % | 0 (0–11.63) | 0 (0–5.36) | 0.932 |
| The percentage of each shape and dimension abnormality was calculated by dividing the sum of the capillaries with that morphology or dimension abnormality by the total number of capillaries in the distal row evaluated per image. | |||
Data expressed as mean±SD/median (range) or frequency (%). The p-value was significant if <0.05.
Bold font indicates statistically significant values.
Other parameters, such as abnormal shapes (bushy, bizarre-shaped, and ramified capillaries), dilated capillaries, and giant capillaries, showed no statistically significant differences between groups.
Correlation analysis revealed no significant relation between capillary density and clinical as well as laboratory variables, such as age, disease duration, ESR, CRP, hemoglobin levels, serum albumin, or Mayo disease activity score, with all p-values exceeding 0.05.
When comparing IBD patients with normal capillary density (≥7/mm) to those with low capillary density (<7/mm), no significant differences were observed in terms of age, disease duration, or treatment type. However, RP was significantly more prevalent in the low capillary density group (38.46%) than the normal density group (10.81%), with a p-value of 0.04 as illustrated in Table 4. In the univariate regression analysis, RP was the only significant predictor of low capillary density (OR=5.156, 95% CI=1.122–23.691, p=0.035), whereas age, gender, disease duration, presence of EIMs, serum albumin, hemoglobin, ESR, CRP, disease activity, corticosteroids intake and biologic therapy did not show a significant link (p>0.1). Applying a multivariate model including RP adjusted with age and gender, RP remained a significant predictor of low capillary density (OR=5.808, 95% CI=1.199–28.134, p=0.029). No significant predictors for microhemorrhages were detected by regression analysis testing the previously mentioned variables.
Comparison between patients with normal capillary density and those with low capillary density.
| Variable | Normal mean capillary density (≥7/mm)n=37 | Low mean capillary density (<7/mm)n=13 | p-Value |
|---|---|---|---|
| Age (years) | 34.3±10.63 | 32.77±8.76 | 0.614 |
| Disease duration (years) | 6 (2–15) | 3 (1–11) | 0.143 |
| Sex | Male: 11 (29.73%) | 5 (38.46%) | 0.4 |
| Female: 26 (70.27%) | 8 (61.54%) | ||
| Raynaud's phenomenon | Yes: 4 (10.81%) | 5 (38.46%) | 0.04 |
| No: 33 (89.19%) | 8 (61.54%) | ||
| Extraintestinal manifestations | Yes: 4 (10.81%) | 2 (15.38%) | 0.497 |
| No: 33 (89.19%) | 11 (84.62%) | ||
| Biologic therapy | Yes: 12 (32.43%) | 5 (38.46%) | 0.741 |
| No: 25 (67.57%) | 8 (61.54%) | ||
| Corticosteroids | Yes: 9 (24.32%) | 2 (15.38%) | 0.7 |
| No: 28 (75. 68%) | 11 (64.62%) | ||
| Mesalazine | Yes: 25 (67.57%) | 8 (61.53%) | 0.741 |
| No: 12 (32.43%) | 5 (38.46%) | ||
| Azathioprine | Yes: 25 (67.57%) | 8 (61.53% | 0.741 |
| No: 12 (32.43%) | 5 (38.46%) | ||
| ESR (mm/h) | 25 (11–60) | 19 (5–51) | 0.418 |
| CRP (mg/dl) | 12 (1–50) | 9 (1–51) | 0.808 |
| Hemoglobin (g/dl) | 10.77±0.9 | 10.69±0.95 | 0.806 |
| Serum albumin (g/dl) | 3.38±0.54 | 3.51±0.59 | 0.227 |
Data expressed as mean±SD/median (range) or frequency (%). The p-value was significant if <0.05. ESR: erythrocyte sedimentation rate, CRP: C-reactive protein.
The NFC standardization and indications gradually expanded in systemic autoimmune diseases.19 Vascular endothelium is involved in the pathogenesis of inflammatory rheumatic diseases such as RA, SLE, SSc, and systemic vasculitis.20 IBD causes persistent intestinal inflammation and affects systemic health21 and vascular endothelium.5 This study used NFC to assess capillary density, shape, dimensions and microhemorrhage frequency in IBD patients compared to a control group, thereby evaluating microcirculation involvement in these individuals.
To our knowledge, this is the first study to explore NFC abnormalities in adult UC patients. Few studies have examined the link between microangiopathy and IBD since its original linkage with UC in 1949.22 A systematic literature review concluded that different types of vasculitis can occur in IBD patients. The diagnosis of IBD can preceded that of vasculitis by years.23
In the current study, IBD patients had significantly reduced capillary density (7.52±0.68capillaries/mm) compared to the control group (8.18±0.63capillaries/mm) (p<0.001). In a previous study conducted in adults, Gasser et al.10 observed an average capillary density of 5.5/mm in CD patients versus 7.1/mm in controls (p=0.001), which is consistent with our results. Kurowski et al.24 reported reduced capillary density in pediatric IBD patients compared to controls (p-value<0.001), which aligns with our results. A study conducted in children observed that capillary density, length, and width increase with age.25 However, in children aged >10 years, the capillary density is similar to that of adults.26 Therefore, we compared p-values of our study with those of Kurowski et al., not the absolute values.
In IBD patients, abnormal capillary morphologies were rare and exhibited no appreciable variation from controls (p>0.05). Moreover, dilated and giant capillaries did not show significant variations between IBD patients and controls (p=0.683 and 0.932, respectively). Studies on SSc, which reflect direct vascular involvement, showed that morphological markers of microvascular damage characterize >95% of patients with overt SSc even if they are not detected concurrently.27 Although IBD influences capillary density and integrity, it does not cause significant structural damage that distinguishes diseases with primary vascular pathology.
Age and disease duration showed a nonsignificant correlation with capillary density. Likewise, disease activity and markers of systemic inflammation showed no substantial relationship with capillary density. This absence of correlations in our study may reflect that capillary alterations represent chronic vascular remodeling rather than acute inflammatory status, which is not captured in a cross-sectional design. Kurowski et al.24 reported a non significant link between capillary density and age, gender, ESR, CRP and disease activity in pediatric IBD patients. However, they observed a significant association between changes in capillary density and disease activity in seven CD patients over 6-month, p=0.047, highlighting the role of the longitudinal assessment.
In our analysis, biologic treatment showed an insignificant relation with capillary density and microhemorrhage frequency. Kurowski et al.24 revealed that capillary density in pediatric IBD patients was generally independent of therapy type, suggesting that microvascular changes may persist after biologic intervention.
Another important finding of our study was that only RP frequency differed between IBD patients with normal (≥7capillaries/mm) and poor (<7capillaries/mm) capillary density. RP was more common in patients with low capillary density (p=0.04) than those with normal density (38.46% vs. 10.81%, respectively). This finding is supported by the results of the regression analysis. Lower capillary density may be related to higher RP sensitivity in IBD patients, indicating a more widespread pattern of microvascular dysfunction in persons with reduced capillary integrity. Age, disease duration, gender distribution, and therapy type did not differ between groups. Low capillary density was unrelated to disease duration (p=0.143), suggesting that low capillary density may not necessarily correlate with IBD disease progression. In accordance of these results, Kurowski et al.24 reported that capillary density did not significantly differ when comparing age and gender.
EIMs of IBD incorporate a wide range of organ system involvement, such as skin manifestations, arthritis, spondylitis, and uveitis.3 In our cohort, 12% of the patients had EIMs. However, when comparing patients with low versus normal capillary density, there was no significant difference in the prevalence of EIMs. Previous studies demonstrated reduced capillary density in uveitis patients (p<0.001),28,29 and capillary morphological alterations without significant capillary loss in ankylosing spondylitis patients,8 when compared with controls. Given the small number of patients with EIMs in our study, a definitive link with NFC changes cannot be established.
The strength of our study is that it is one of the few to investigate NFC findings in adult IBD patients, addressing a significant knowledge gap in understanding microvascular involvement in this population. The identification of association between NFC results, such as reduced capillary density with RP, adds valuable insights into IBD EIMs. Our study's limitations include a relatively small sample size and small number of CD patients which limits the ability of comparative analysis between IBD subtypes and reduces the generalizability of the results, particularly for CD patients. However, the small number of CD patients reflects the real-life predominance of UC over CD in our country.30 Our study focused on a single center, which may not have captured variability in patient characteristics and therapeutic practices. The identification of patients with RP was based on a clinical approach without applying standardized confirmatory tests. The risk of selection bias was minimized by applying strict exclusion criteria for both groups and matching cases and controls for age, gender and BMI. However, the study included preliminary findings that require further investigation, with a longitudinal component and larger sample size.
Conclusion and recommendationsCompared to healthy controls, reduced capillary density and increased microhemorrhages in IBD patients suggested systemic microvascular involvement independent of disease duration or inflammatory markers. Although it did not correlate with systemic inflammatory markers, NFC could be a useful non-invasive method for evaluating microvascular alterations in IBD patients. Future research with larger, multicenter cohorts involving multiple observers and longitudinal approaches is recommended to confirm these results, explore their implications for disease monitoring and formally assess reproducibility and strengthen external validity.
Authors’ contributionsThe study conception and design were done by Dr Maha S.I. Abdelrahman. All authors contributed to the acquisition of data. Dr. Maha S.I. Abdelrahman was responsible for data analysis and interpretation. All authors participated in drafting the article and reviewing it critically for important intellectual aspects, and all authors approved the final version to be published.
Consent for publicationNot applicable.
FundingThis study did not receive any financial support.
Conflict of interestThe authors declare that they have no conflict of interest.
AcknowledgementsAssiut Tropical IBD Study Group (ATISG).
Data availabilityThe present study's data are available from the corresponding author upon request.
