Clinical Anatomy 21:314–318 (2008)

ORIGINAL COMMUNICATION

Clubbed Fingers: Radiological Evaluation

of the Nail Bed Thickness
ANA-LUIZA S. MOREIRA, NELSON S. PORTO, JOSÉ S. MOREIRA,
JANE M. ULBRICH-KULCZYNSKI, AND KLAUS L. IRION*
The Post-Graduation Program in Chest Medicine Sciences of the Federal University of Rio Grande do Sul,
Porto Alegre, Brazil

It is established that there is an increase in soft tissue of the clubbed digits as

demonstrated on previous histopathological examinations. In the present
study, the nail bed thickness was assessed and measured on plain radiographs
of index fingers in two groups of patients: one group with lung disease and fin-
gers clubbing and one group of normal controls. A vertical x-ray beam was
used with a focus-film distance of 1.0 m, with the index finger placed in lateral
view directly over the film, without anti-diffusion grid. Three investigators,
blinded to prevent bias measured the thickness of soft tissues between the
nail root and the terminal phalanx on the radiographs. This method was used
to evaluate a group of 85 clinically clubbed (hyponychial angle > 192.08) adult
patients with lung disease and a control group of a 100 normal adult individu-
als with no clubbing (hyponychial angle < 188.08). The mean nail bed thick-
ness in the patients with clubbing (n ¼ 85) was 3.88 6 0.55 mm (3.00–5.50
mm). In comparison, in the normal subjects (n ¼ 100), the mean was 2.38 6
0.27 mm (1.75–3.10 mm), revealing a significant difference (P < 0.001). Only
two normal individuals presented nail bed thickness
3.0 mm. A good interob-
server agreement on the measurements was found (P > 0.900). The radio-
graphic evaluation of the nail bed thickness was easily performed, with good
interobserver concordance. It is possible to distinguish between clubbed from
nonclubbed fingers, in vivo, using plain radiograph. Clin. Anat. 21:314–318,
2008. V 2008 Wiley-Liss, Inc.
C

Key words: finger; hyponychial angle; clubbing, paraneoplasic syndrome,

hypocratism; nail; lung cancer; x-rays

INTRODUCTION thoracic diseases, as studied by Epstein et al.

(1979), Kitis et al. (1979), and Fatourechi et al.
Finger clubbing is an important clinical sign known (2002); and in hereditary diseases, as demonstrated
since the time of Hippocrates, as cited by Major by Talbott and Montgomery Jr. (1953), Curth et al.
(1945). The abnormality is characterized by an (1961) and Fischer et al. (1964). Clubbing can
increase in the volume of the soft tissue in the nail appear as an isolated finding or can be part of the
beds of fingers with consequent obliteration of the
angle formed between the nail and the dorsal aspect
of the skin of the distal phalanx, as studied by Trous- *Correspondence to: Klaus L. Irion, Department of Radiology,
seau (1834) and Lovibond (1938). This clinical sign The Cardiothoracic Centre Liverpool NHS Trust, Thomas Drive,
has been reported to occur in a variety of acquired Liverpool, United Kingdom. E-mail: [email protected]
diseases: in pulmonary diseases, as demonstrated Received 14 May 2007; Revised 25 February 2008; Accepted
by Dickinson (1992) and Sridhar et al. (1998); in 29 February 2008
cardiac diseases, as demonstrated by Diop et al. Published online 18 April 2008 in Wiley InterScience (www.
(1997) and Koegelenberg et al. (2003); in extra- interscience.wiley.com). DOI 10.1002/ca.20623

C 2008
V Wiley-Liss, Inc.
 Radiology of Clubbed Fingers 315

Fig. 1. A: Index finger

positioning for x-rays expo-
sure. The vertical projection
of the x-ray beam with
focus-film distance fixed at
1.0 m. B: True lateral view
of the index finger, with the
finger in contact with the
film cassette (in order to
minimize magnification) and
without anti-diffusion grid to
(reduce the dose).

Bamberger-Marie complex, as published by Bam- patients with cystic fibrosis with clubbing. Ward
berger (1889) and by Marie (1890) and can be part of et al. (1995) documented a more intense glucose
the hypertrophic osteoarthropathy (HOA) syndrome, metabolism in the fingertips of a patient with lung
as demonstrated by Mendlowitz (1942) and Stenseth cancer with clubbing using positron emission tomog-
et al. (1967). Generally clubbing is symmetric occur- raphy.
ring in both hands and both feet, but it can also mani- Our study compared clubbed and normal fingers
fest asymmetrically and unilaterally, as described by using plain radiographs. The aim was to objectively
Kaditis et al. (1995) and Abe et al. (1999). document the measurements of the nail bed thick-
Accordingly to Myers and Farquhar (2001), the di- ness (NBT) of clubbed and of normal fingers in living
agnosis of clubbing is clinical, but it may be corrobo- patients, probably the most significant anatomical
rated by some measurements, such as the hypony- change responsible for clubbing.
chial angle and the distal phalangeal/interphalangeal
depth ratio (DPD/IPD). Hyponychial Angle >192.08
and DPD/IPD ratio >1 indicate the presence of finger METHODS
clubbing, as demonstrated by Regan et al. (1967),
Waring et al. (1971), Husarik et al. (2002), and Lateral x-ray projections of the right index finger
Moreira et al. (2004). positioned directly over the film cassette without
The pathogenesis of clubbing is not yet fully anti-scatter grid (Fig. 1) were obtained from each of
understood and the theoretical explanations were the 85 patients with clubbing. Of these 85 patients,
explored by Shneerson (1981), Dickinson and Martin 63 were male and 22 were female, aged between 15
(1987), and by Atkinson and Fox (2004). They have and 80 years old (51.8 6 13.4 years). All these
postulated that growth factors, especially cytokines patients clinically presented with clubbed fingers as
related to megakaryocytes, may be involved in the a consequence of lung diseases such as carcinoma,
process and can accumulate at sites of the abnor- tuberculosis, idiopathic pulmonary fibrosis, bron-
malities. chiectasis, and lung abscess. Plain film images using
Post mortem histopathological studies of clubbed the above-described technique were also performed
fingers were carried out by Pigeaux (1832), Locke on the control group of 100 individuals of which 59
(1915), Crump (1929), Lovell (1950), Gall et al. were male and 41 were female, mean age 39.1 6
(1951), Bigler (1958), Turner-Warwick (1963), Pon- 14.0 years old (16–74 years old). None of the con-
chon et al. (1969), and Currie and Gallagher (1988). trols had any clinical symptoms of chest disease or
All these authors found increased soft tissue in the clinical evidence of fingers clubbing. Eighty-seven of
fingertips of clubbed fingers. Some of these studies them (87.0%) were nonsmokers, and 81 (81.0%)
used angiograms, and documented local vascular in- had a normal chest x-rays in the previous 5 years.
crement in some cases. On each film (Fig. 2), three independent observers
Bigler (1958) has demonstrated by histopathology blind to the clinical findings traced a line to measure
that, only in cases of clubbing, the thumb nail bed the nail bed thickness corresponding to the soft tis-
thickness was greater than 2.0 mm. More recently, sue thickness interposed between the nail base and
Rush et al. (1992) using thermography found a rise the dorsal cortex of the distal phalanx, using a 0.05-
of temperature by a half a degree in fingers of mm Vernier caliper. Subsequently, the hyponychial
316 Moreira et al.

Fig. 2. Measurement of the nail bed thickness (ver- nail bed). A: Measurements and determinations in three
tical line perpendicular to the distal phalanx, at the nail clubbed finger patients. B: Measurements and determi-
bed) and hyponychial angle (angle formed by the two nations in three normal patients.
lines longitudinal to the index finger, with vertex at the

angle was measured using a half degree calibrated Informed consent was obtained from all patients
protractor. and controls, and the Post-Graduation Program
The clinical diagnosis of clubbing was made when Committee approved the study.
the hyponychial angle was greater than 192.08 (Fig. Statistical tests (T-test, ANOVA) were used to
2). The diagnosis of absence of clubbing was made assess the interobserver agreement for the values of
when the hyponychial angle was less than 188.08 the hyponychial angle and of the nail bed thickness
(Fig. 2). measurements, and also to verify differences

TABLE 1. Nail Bed Thickness Measurements and Hyponychial Angle Values in Clubbed and in
Nonclubbed Fingers Verified by 3 Investigators in Clubbed Patients and in Control Normal Individuals
Clubbed Fingers (85 patients) Nonclubbed Fingers (100 normal controls) P
Nail bed thickness
Investigator 1 3.89 6 0.55 mm 2.39 6 0.25 mm <0.001
(3.00–5.50) (1.80–3.00)
Investigator 2 3.88 6 0.54 mm 2.38 6 0.27 mm <0.001
(3.00–5.50) (1.75–3.00)
Investigator 3 3.88 6 0.55 mm 2.40 6 0.26 mm <0.001
(3.00–5.50) (1.75–3.10)
Hyponychial angle
Investigator 1 198.8 6 5.28 179.8 6 4.68 <0.001
(192.0–222.0) (172.5–187.0)
Investigator 2 198.5 6 5.28 180.0 6 4.78 <0.001
(192.0–222.0) (172.5–187.0)
Investigator 3 199.0 6 5.18 179.7 6 4.68 <0.001
(192.0–222.5) (172.5–187.0)

Very strong interobserver agreement was observed: in nail bed thickness (P ¼ 0.998 in clubbed; P ¼ 0.979 in non-
clubbed fingers); and in hyponychial angles (P ¼ 0.997 in clubbed; P ¼ 0.978 in nonclubbed fingers). Significant
differences between the two groups (with clubbing and normal controls).
 Radiology of Clubbed Fingers 317

between the clubbed and normal fingers. Pearson’s

correlation test was used to compare the nail bed
thickness and the hyponychial angle values. The
adopted significance level was 0.05.

RESULTS
Each of the three investigators performed meas-
urements on each radiograph from 85 clubbed fin-
gers and 100 nonclubbed fingers, resulting in a total
of 255 measurements of nail bed thickness and
determinations of the hyponychial angle on clubbed
fingers to compared with the total of 300 measure-
ments and determinations from the normal control
fingers. Statistically significant differences were
found. The nail bed thickness measured 3.88 6 0.55
mm and the hyponychial angle measured 198.8 6
5.28 (range ¼ 192.0–222.08) in the 85 subjects pre- Fig. 4. Correlation between hyponychial angle (HA)
senting with clubbed fingers. In the control group and nail bed thickness (NBT) verified in the whole group
the same parameters measured 2.38 6 0.27 mm of 185 controls and patients (r ¼ 0.870).
and 180.1 6 3.78 (range ¼ 170.0–188.08), respec-
tively (Table 1 and Fig. 3). Both differences were
statistically significant (P < 0.001). No significant nail bed thickness measured 2.43 6 0.28 mm in the
intra or interobserver differences were found in the normal males, and 2.31 6 0.28 mm in the normal
nail bed thickness measurements performed by each females (P ¼ 0.058). The hyponychial angle meas-
of the three investigators in both the clubbed (P ¼ ured 179.4 6 3.98 and 180.7 6 3.38 (P ¼ 0.111) in
0.998) and in the normal (P ¼ 0.979) groups. The normal males and normal females, respectively.
nail bed thickness was
3.00 mm in all the 85 Considering all individuals (N ¼ 185), the correlation
clubbed fingers and was <3.00 mm in 98 subjects of coefficient (Pearson) between the nail bed thickness
the 100 individuals in the control group. Two of the and hyponychial angle values was 0.870 (Fig. 4).
normal controls broke the rule and had nail bed
thickness of 3.00-mm and 3.10-mm, respectively.
Additionally, the influence of age was studied in
the control group. The nail bed thickness measured DISCUSSION
2.33 6 0.26 mm in the normal individuals aged 40 or
younger and measured 2.44 6 0.32 mm in the nor- The radiographic images of the index fingers were
mal subjects who were over 40 years old. This differ- easily obtained with virtually no projectional distor-
ence was not statistically significant (P ¼ 0.916). In tion since the digits were in direct contact with the
the patients with clubbed fingers, the hyponychial film cassette with a fixed focus-film distance of 1.0
angle measured 180.2 6 3.08 in the sub-group m. No major magnification is expected when the
younger than 40, and 180.0 6 4.18 (P ¼ 0.799) in object (finger) is in contact with the film cassette.
the subgroup over 40 years, again not a statistically Based on this we considered the measured values as
significant difference. With respect to gender, the the true values of the individuals’ digits. The other
advantage of obtaining the film with the finger in
contact with the cassette is that this avoids image
blurring, thus providing well-demarcated landmarks
for the measurements, which was reflected by the
absence of interobserver variability.
In order to make the meaning of the ‘‘in-vivo’’
more relevant, the values of the nail bed thickness
were considered only from the patients with clinically
diagnosed fingers clubbing and confirmed by the
presence of abnormal hyponychial angle. None in the
control group had clinical evidence of finger clubbing,
all of them had hyponychial angle within normal
limits, as established by Regan et al. (1967) and
Husarik et al. (2002), and the majority of them were
nonsmokers.
Previous verifications studying healthy subjects
revealed that the hyponychial angle was not signifi-
cantly different between genders in a series of 123
Fig. 3. Comparative values of the nail bed thickness studied by Husarik et al. (2002) and another series
(NBT) in 100 normal controls and in 85 patients with of 452 studied by Moreira et al. (2004). Our study
clubbing. Significant difference (P < 0.001). revealed no significant age related differences in the
318 Moreira et al.

observed nail bed thickness of normal subjects. Fischer DS, Singer DH, Feldman SM. 1964. Clubbing, a review, with
However, marginal difference (P ¼ 0.058) was found emphasis on hereditary acropachy. Medicine 43:459–479.
when comparing genders indicating possible smaller Gall EA, Bennett GA, Bauer W. 1951. Generalized hypertrophic
osteoarthropathy. Am J Pathol 27:349–381.
values in females of this group.
Husarik D, Vavricka SR, Mark M, Schaffner A, Walter RB. 2002.
The present study also revealed a significant Assessment of digital clubbing in medical inpatients by digital
increase in nail bed thickness in clubbed fingers from photography and computerized analysis. Swiss Med Wkly 132:
affected patients compared to individuals from the 132–138.
normal control group. These finding are in concord- Kaditis AG, Nelson AM, Driscoll DJ. 1995. Takayasu’s arteritis present-
ance to what had been reported by other authors ing with unilateral digital clubbing. J Rheumatol 22:2346–2348.
when studying histopathology specimens, including Kitis G, Thompson H, Allan RN. 1979. Finger clubbing in inflamma-
Lovell (1950), Bigler (1958), and Currie and Gallh- tory bowel disease: Its prevalence and pathogenesis. Br Med J
ager (1988). These studies also documented the dis- 2:825–828.
tribution pattern of the soft tissue thickening in fin- Koegelenberg CF, Doubell AF, Orth H, Reuter H. 2003. Infective en-
docarditis in the Western Cape Province of South Africa: A
gertips, particularly in the nail bed of patients with three-year prospective study. QJM 96:217–225.
clubbing as quantitatively analyzed by Bigler (1958). Locke EA. 1915. Secondary hypertrophic osteo-arthropathy and its
Finally an excellent degree of correlation has been relation to simple club-fingers. Arch Intern Med 15:660–713.
demonstrated, between the nail bed thickness and Lovell RRH. 1950. Observations on the structure of clubbed fingers.
hyponychial angle values. This suggests a directly Clin Sci 9:299–321.
proportional relation of the above mentioned meas- Lovibond JL. 1938. Diagnosis of clubbed fingers. Lancet 1:363–
urements. 364.
Major RH. 1945. Hippocrates. The Book of prognostics. In: Classic
Descriptions of Disease. 3rd Ed. Springfield: Charles C Thomas.
ACKNOWLEDGMENT p 4–5.
Marie P. 1890. De l’ostéo-arthropathie hypertrophiante pneumique.
The authors thank Dr. Ravi Kodur for his help with Rev Méd (Paris) 10:1–36.
the manuscript. Mendlowitz M. 1942. Clubbing and hypertrophic osteoarthropathy.
Medicine 21:269–306.
Moreira da Silva J, Porto Nda S, Moreira A. 2004. Objective evalua-
tion of clubbing on shadow images of index fingers. A study of
REFERENCES patients with pulmonary disease and of normal individuals.
Abe K, Sasaki T, Kawazoe K, Nakamura S. 1999. Pal mar aneurysm J Bras Pneumol 30:36–44.
accompanied by ipsilateal clubbing finger—A case report. Angiol- Myers KA, Farquhar DR. 2001. The rational clinical examination.
ogy 50:69–73. Does this patient have clubbing? JAMA 286:341–347.
Atkinson S, Fox SB. 2004. Vascular endothelial growth factor Pigeaux J. 1832. Recherches nouvelles sur l’étiologie, la symptoma-
(VEGF)-A and platelet-derived growth factor (PDGF) play a cen- logie et le mécanisme du developpement fusiforme de
tral role in the pathogenesis of digital clubbing. J Pathol l’extremité des doigts. Arch Gén Méd 29:174–184.
203:721–728. Ponchon Y, Chelloul N, Roujeau J. 1969. Anatomo-pathological
Bamberger E. 1889. Sitzungsbericht der KK. Gesellschaft der Ärzte study of digital hippocratism. Sem Hop 45:2604–2611.
in Wien von 8. März 1889. Wien Klin Wochenschr 2:225–226. Regan GM, Tagg B, Thomson ML. 1967. Subjective assessment and
Bigler FC. 1958. The morphology of clubbing. Am J Pathol 34:237–261. objective measurement of finger clubbing. Lancet 1:530–532.
Crump C. 1929. Histologie der allgemeinen osteophytose (Ostéo- Rush PJ, Giorshev C, Shore A, Levinson H. 1992. The use of ther-
artrhopathie hypertrophiante pneumique). Virchows Arch f Path mography in clubbing. Respir Med 86:257–259.
Anat 271:467–511. Shneerson JM. 1981. Digital clubbing and hypertrophic osteoartr-
Currie AE, Gallagher PJ. 1988. The pathology of clubbing: Vascular opathy: The underlying mechanisms. Br J Dis Chest 75:113–
changes in the nail bed. Br J Dis Chest 82:382–385. 131.
Curth HO, Firschein IL, Alpert M. 1961. Familial clubbed fingers. Sridhar KS, Lobo CF, Altman RD. 1998. Digital clubbing and lung
Arch Dermatol 83:828–836. cancer. Chest 114:1535–1537.
Dickinson CJ, Martin JF. 1987. Megakaryocytes and platelet clumps Stenseth JH, Clagett OT, Woolner LB. 1967. Hypertrophic pulmo-
as the cause of finger clubbing. Lancet 2:1434–1435. nary osteoarthropathy. Dis Chest 52:62–68.
Dickinson CJ. 1992. Lung diseases associated with digital clubbing. Talbott JH, Montgomery WR Jr. 1953. Familial clubbing of fingers
Clin Exp Rheumatol (10 Suppl) 7:23–25. and toes. AMA Arch Intern Med 92:697–700.
Diop IB, Sarr M, Kane A, Hane L, Dieye O, Ndiaye M, Ba K, Sow D, Trousseau A. 1834. De la forme hipocratique des doigts de tubercu-
Fall M, Diouf SM. 1997. Tetralogy of Fallot. Anatomo-clinical, leux. J Conaiss Méd Chir 1:351–352.
prognostic and therapeutic features. Dakar Med 42:118–122. Turner-Warwick M. 1963. Systemic arterial patterns in the lung and
Epstein O, Ajdukiewicz AB, Dick R, Sherlock S. 1979. Hypertrophic clubbing of the fingers. Thorax 18:238–250.
hepatic osteoarthropathy. Clinical, roentgenologic, biochemical, Ward RW, Chin R Jr, Keyes JW Jr, Haponik EF. 1995. Digital club-
hormonal and cardiorespiratory studies, and review of the litera- bing. Demonstration with positron emission tomography. Chest
ture. Am J Med 67:88–97. 107:1172–1173.
Fatourechi V, Ahmed DD, Schwartz KM. 2002. Thyroid acropachy: Waring WW, Wilkinson RW, Wiebe RA, Faul BC, Hilman BC. 1971.
Report of 40 patients treated at a single institution in a 26-year Quantitation of digital clubbing in children. Measurements of
period. J Clin Endocrinol Metab 87:5435–5441. casts of the index finger. Am Rev Respir Dis 104:166–174.