83
DERMATOGLYPHIC STUDY OF SYSTEMIC LUPUS ERYTHEMATOSUS ROBERT W. DUBOIS, JOHN M. WEINER, and EDMUND L. DUBOIS Dermatogl yphic patterns were analyzed from two races of patients with SLE: a Mexican-American series of 27 females with SLE and 28 matched controls, and a Caucasian series of 28 females with SLE and 26 matched controls. Eighty-five measurements and 23 indices were analyzed. From these data, eleven parameters were statistically significant, four separating the Mexican-American SLE group from their controls and seven separating the Caucasian SLE group from their controls (P values < 0.002 to < 0.05). Only two parameters were significantly different between the two normal series, but nine parameters differentiated the Mexican-American SLE from the Caucasian SLE groups. A multiple linear discriminant function was computed using the most significant parameters. A misclassification rate of 25-30% was obFrom the Section of Clinical Immunology and Rheumatic Disease, Department of Medicine, University of Southern California School of Medicine, Los Angeles, California. Supported by USPH Grant AM-15755. Robert W. Dubois: Student Fellow in Rheumatology, Section of Clinical Immunology and Rheumatic Disease, Department of Medicine, University of Southern California School of Medicine; John M. Weiner. Dr.P.H.: Associate Professor of Medicine and Chief, Clinical Research Information Systems Unit, Department of Medicine, University of Southern California School of Medicine; Edmund L. Dubois, M.D.: Clinical Professor of Medicine, Section of Clinical Immunology and Rheumatic Disease, Department of Medicine, University of Southern California School of Medicine. Address reprint requests t o Edmund L. Dubois. M.D., Department of Medicine, University of Southern California School of Medicine, 2025 Zonal Avenue, Los Angeles, California 90033. Submitted for publication March 27, 1975; accepted June 30, 1975. Arthritis and Rheumatism, Vol. 19, No. 1 (January-February 1976)
served between patients and matched controls. Dermatoglyphic patterns have been shown to be genetically controlled. These data suggest a genetic abnormality associated with SLE. Dermatoglyphics, as defined by Cummins and Midlo, is the study of ridged skin that appears on the fingers, toes, palms, and soles (1 ). Although the exact methodology of fingerprint pattern inheritance is unknown, it has been shown to be genetically controlled in sibling studies and in chromosomal abnormalities (2). Rose et a1 in 1972 reported dermatoglyphic differences in women with increased fetal wastage (3). These findings stimulated research in women with SLE, because in this disease it has been shown that the incidence of spontaneous abortions is higher than in a control group (4).Three studies to date have investigated the dermatoglyphic patterns of patients with SLE. Dubes, Fraga, and Qazi all observed statistically significant findings between SLE and matched control populations (5-7). However in each of the three studies the significant features were different. These conflicting findings stimulated this current study which uses a larger population and covers more parameters than the previous work, in order to determine if any significant difference exists.
MATERIALS AND METHODS Finger and palm prints were obtained from two groups of SLE patients and compared with two sets of age-, race-, and sex-matched controls. The patients fulfilled the preliminary
DUBOIS ET A L
84
IGIT RIDGE COUNT,
FI NGER LENG
a-b RIDGE
COUNT.
PALMAR PATTERNS
t-a DISTANCE
ULNARWRIST CREASE (WC)
t - TRlRADUlS LATERAL DISPLACEMENT Fig I . Dertnatoglyphic structure.
A R A criteria for the diagnosis of SLE (8). One group consisted of 27 Mexican-American females (characterized by Spanish surnames and racial characteristics) with SLE who were compared with 28 controls and the other group was 28 Caucasian females with SLE compared with 26 controls. The normal females had no past or family history of rheumatic disease and were either attending obstetric clinics or volunteers. The authors used the FaurotB process of inkless prints on sensitized paper for recording the dermatoglyphic patterns (Faurot, Inc, New York, New York). Becauseof Rose's observation on fetal wastage (3), the subjects were questioned concerning the number of spontaneous abortions and successful pregnancies. The following measurements and data were compiled: finger patterns and ridge counts on all 10 fingers, palmar patterns in the hypothenar, thenar first, second, third, and fourth interdigital areas, a-b ridge counts, axial t-triradii placement, finger length, and atd angles (Figure I ) . The axial t-triradius vertical placement was measured by dropping a perpendicular line from the t-triradius to a line approximating the wrist crease. Its lateral displacement was determined by the distances between the t-triradius and perpendicular lines dropped from the base of the index finger and the ulnar edge of the hand to the wrist crease. Measurements were rechecked blind and all statistically significant features were remeasured independently by two observers. The data obtained were statistically analyzed using Student's t and chi-square test where applicable. Later the
most significant parameters were analyzed using multiple linear discriminant analysis (9).
RESULTS Eleven features proved significant out of the 85 original measurements and 23 calculated indices examined in either the Mexican-American female or the Caucasian female groups. In separating t h e Mexican-American normals from the SLEs, four parameters were statistically significant (Figure 2). T h e radial measurement of the axial t-triradius in the right hand of the Mexican-American S L E group was decreased when compared with the control ( P < 0.0027). The composite measurement of the right axial t-triradius (radial measurement/ulnar measurement) also showed a significant difference ( P < 0.0036). Both these results indicate a radial shift of the t-triradius in S L E patients. In the left hypothenar area, more patterns appeared in the Mexican-American S L E group ( P < 0.036). In the left-hand first digit (thumb) in this group, more whorls were observed than in the control group where more loops were observed ( P < 0.002). In the Caucasian SLE group, seven significant parameters were observed (Figure 3). In the left hand of the SLE group, both the ulnar displacement of the left axial t-triradius ( P < 0.0026) and the composite mea-
DERMATOGLYPHIC STUDY OF SLE
85
pattern in the left hypothenar area, and vertical displacement of the right axial t-triradius (Table 2). The misclassification rate was 26% for the 54 in this group ( P < 0.0012).
DISCUSSION
LEFT
RIGHT
t-TRIR ADllJS SPLACED RADIALLY
\WRIST
The three previous investigations of dermatoglyphics and SLE showed different significant features. Dubes observed a displacement of the axial t-triradius (5). Fraga and Qazi found changes in the pattern frequencies of the fingers (6, 7). Definite racial differences have previously been shown in dermatoglyphic patterns (10, 11). Therefore, because Fraga observed Mexicans, Dubes observed Caucasians, and Qazi observed blacks, their data should be different. The authors of this paper analyzed both Mexican-Americans and Caucasians, and the results in the Caucasian SLE group agree with those of Dubes. He observed an ulnar displacement of the right axial t-triradius, and the present data concur ( P < 0.05). Dubes also found an ulnar displacement of the same measurement predominantly
CREASE
Fig 2. Signijicant parameters in Mexican-A merican groups.
surement (radial displacement/ulnar displacement, P < 0.01 1 ) of the left axial t-triradius were significant, indicating an ulnar shift of the triradius in this group. I n this same hand the SLE group showed a decrease in the log of the a-b ridge count ( P < 0.05) and an increase in hypothenar patterns ( P < 0.036). In the right hand of the SLE group, the authors observed a decrease in the atd angle ( P < 0.042), a decrease in the log of the a-b distance ( P < 0.05), and a proximal displacement of the axial t-triradius ( P < 0.05). The observed number of habitual aborters with SLE was insufficient for analysis. Discriminant analysis was performed using data for each race. In the Mexican-American SLE group the radial measurement of the right axial t-triradius and the presence of a pattern in the left hypothenar area were the only significant parameters in the discriminant function ( P < 0.0028) (Table I ) . The misclassification rate for the 54 in this group was 28%. There were insufficient data to make these measurements in one case in the normal group. In the Caucasian SLE group three parameters were used in the discriminant function: the ulnar measurement of the left axial t-triradius, presence of a
RlST CREASE
LEFT
RIGHT RIDGE COUNT LESS DISPLACEMENT LE
SLE atd
ANGLE LES
DISPLACED ULNAD
PROXIMALLY DISPLACED’ Fig 3. Significant parameters in Caucasian groups.
DUBOIS ET AL
86
Table 1. Discriminant Analysis of Mexican-Americans Using Two Param ters 1 ) Radial displacement of right axial t-triradius 2) Presence of a pattern in the left hypothenar area ~
Dermatographically Analyzed as: ~~
Diagnosed Normal
Diagnosed SLE
Totdl
20 1
n
28
19
26
21
54
~~~
Normal SLE TOTAL
P
21 * < 0.0028
Misclassification rate: 28%.
* Insufficient data in one case.
in the right hand. In the present series the ulnar shift of this same measurement was statistically more significant in the left hand. However several of Dubes’ other less statistically significant findings were not confirmed in the present study. The discrepancy could be due to sampling error, because his Caucasian SLE group consisted of only 7 females. This study did not deal with black SLE patients. No correlation can be drawn. In the Mexican SLE patients Fragaetal observed a diminution of radial loops and digital arches ( 6 ) . In t h e present series a significant increase in whorls (P < 0.002) was found only in the left-hand first finger. Fraga also observed more patterns i n the thenar first interdigital region and the axial t-triradius in t’. The data of the present authors do not substantiate either of these findings. Possibly, Fraga’s Mexican population differed from the Mexican-American group in the Southern California area. Dermatoglyphic differences are known to exist between races (10, 11). In this investigation two features were observed to be statistically significant in separating the two normal groups: the ridge density on the left thumb ( P < 0.046) and the a-b ridge density on the left
hand (P < 0.003). The same approximate result was expected between the Mexican-American SLE group and the Caucasian SLE group. In fact nine significant parameters differentiated the two SLE groups, including the two stated above. These other features included: atd angle right hand (P < 0.014), radial ( P < 0.0033) and ulnar (P < 0.00091) measur,ement of the right axial ttriradius, ulnar measurement of the left axial t-triradius ( P < 0.0022), pattern types on the right fingers (P < 0.05),and the axial t-triradius lateral displacement (composite measurement) in both the right (P< 0.00021) and left (P < 0.013) hands. The discriminant analysis showed a misclassification rate of 25-30%. Although dermatoglyphic changes were associated with SLE, the measurements used were not sufficiently different to suggest their use in the earlier detection of SLE patients. Several genetically inherited diseases show abnormalities in dermatoglyphics. Differences have been observed in the palm prints of patients with Down’s syndrome and those with Klinefelter’s syndrome when they were compared with matched controls (12). Furthermore finger and palm prints have been shown to be genetically controlled (2). Viral infections such as congenital rubella may also modify dermatoglyphic patterns (1 3, 14). Therefore the finding of statistically significant differences between the SLE and normal groups suggests either a genetic abnormality or alteration by a congenitally transmitted virus or perhaps both as a cause of these changes.
REFERENCES 1 . Cummins H, Midlo C: Palmar and plantar epidermal
2.
3. Table 2 . Discriminant Analysis of Caucasians Using Three Parameters I ) Ulnar displacement of left axial t-triradiur 2 ) Presence of a pattern in the left hypothenar area 3 ) Vertical displacement of right axial t-triradius Dermatographically Analyzed as:
Diagnosed Normal
Diagnosed SLE
Total
Normal SLE
19
7 21
26 28
28
54
1
TOTAL
P
DERMATOGLYPHIC STUDY OF SYSTEMIC LUPUS ERYTHEMATOSUS ROBERT W. DUBOIS, JOHN M. WEINER, and EDMUND L. DUBOIS Dermatogl yphic patterns were analyzed from two races of patients with SLE: a Mexican-American series of 27 females with SLE and 28 matched controls, and a Caucasian series of 28 females with SLE and 26 matched controls. Eighty-five measurements and 23 indices were analyzed. From these data, eleven parameters were statistically significant, four separating the Mexican-American SLE group from their controls and seven separating the Caucasian SLE group from their controls (P values < 0.002 to < 0.05). Only two parameters were significantly different between the two normal series, but nine parameters differentiated the Mexican-American SLE from the Caucasian SLE groups. A multiple linear discriminant function was computed using the most significant parameters. A misclassification rate of 25-30% was obFrom the Section of Clinical Immunology and Rheumatic Disease, Department of Medicine, University of Southern California School of Medicine, Los Angeles, California. Supported by USPH Grant AM-15755. Robert W. Dubois: Student Fellow in Rheumatology, Section of Clinical Immunology and Rheumatic Disease, Department of Medicine, University of Southern California School of Medicine; John M. Weiner. Dr.P.H.: Associate Professor of Medicine and Chief, Clinical Research Information Systems Unit, Department of Medicine, University of Southern California School of Medicine; Edmund L. Dubois, M.D.: Clinical Professor of Medicine, Section of Clinical Immunology and Rheumatic Disease, Department of Medicine, University of Southern California School of Medicine. Address reprint requests t o Edmund L. Dubois. M.D., Department of Medicine, University of Southern California School of Medicine, 2025 Zonal Avenue, Los Angeles, California 90033. Submitted for publication March 27, 1975; accepted June 30, 1975. Arthritis and Rheumatism, Vol. 19, No. 1 (January-February 1976)
served between patients and matched controls. Dermatoglyphic patterns have been shown to be genetically controlled. These data suggest a genetic abnormality associated with SLE. Dermatoglyphics, as defined by Cummins and Midlo, is the study of ridged skin that appears on the fingers, toes, palms, and soles (1 ). Although the exact methodology of fingerprint pattern inheritance is unknown, it has been shown to be genetically controlled in sibling studies and in chromosomal abnormalities (2). Rose et a1 in 1972 reported dermatoglyphic differences in women with increased fetal wastage (3). These findings stimulated research in women with SLE, because in this disease it has been shown that the incidence of spontaneous abortions is higher than in a control group (4).Three studies to date have investigated the dermatoglyphic patterns of patients with SLE. Dubes, Fraga, and Qazi all observed statistically significant findings between SLE and matched control populations (5-7). However in each of the three studies the significant features were different. These conflicting findings stimulated this current study which uses a larger population and covers more parameters than the previous work, in order to determine if any significant difference exists.
MATERIALS AND METHODS Finger and palm prints were obtained from two groups of SLE patients and compared with two sets of age-, race-, and sex-matched controls. The patients fulfilled the preliminary
DUBOIS ET A L
84
IGIT RIDGE COUNT,
FI NGER LENG
a-b RIDGE
COUNT.
PALMAR PATTERNS
t-a DISTANCE
ULNARWRIST CREASE (WC)
t - TRlRADUlS LATERAL DISPLACEMENT Fig I . Dertnatoglyphic structure.
A R A criteria for the diagnosis of SLE (8). One group consisted of 27 Mexican-American females (characterized by Spanish surnames and racial characteristics) with SLE who were compared with 28 controls and the other group was 28 Caucasian females with SLE compared with 26 controls. The normal females had no past or family history of rheumatic disease and were either attending obstetric clinics or volunteers. The authors used the FaurotB process of inkless prints on sensitized paper for recording the dermatoglyphic patterns (Faurot, Inc, New York, New York). Becauseof Rose's observation on fetal wastage (3), the subjects were questioned concerning the number of spontaneous abortions and successful pregnancies. The following measurements and data were compiled: finger patterns and ridge counts on all 10 fingers, palmar patterns in the hypothenar, thenar first, second, third, and fourth interdigital areas, a-b ridge counts, axial t-triradii placement, finger length, and atd angles (Figure I ) . The axial t-triradius vertical placement was measured by dropping a perpendicular line from the t-triradius to a line approximating the wrist crease. Its lateral displacement was determined by the distances between the t-triradius and perpendicular lines dropped from the base of the index finger and the ulnar edge of the hand to the wrist crease. Measurements were rechecked blind and all statistically significant features were remeasured independently by two observers. The data obtained were statistically analyzed using Student's t and chi-square test where applicable. Later the
most significant parameters were analyzed using multiple linear discriminant analysis (9).
RESULTS Eleven features proved significant out of the 85 original measurements and 23 calculated indices examined in either the Mexican-American female or the Caucasian female groups. In separating t h e Mexican-American normals from the SLEs, four parameters were statistically significant (Figure 2). T h e radial measurement of the axial t-triradius in the right hand of the Mexican-American S L E group was decreased when compared with the control ( P < 0.0027). The composite measurement of the right axial t-triradius (radial measurement/ulnar measurement) also showed a significant difference ( P < 0.0036). Both these results indicate a radial shift of the t-triradius in S L E patients. In the left hypothenar area, more patterns appeared in the Mexican-American S L E group ( P < 0.036). In the left-hand first digit (thumb) in this group, more whorls were observed than in the control group where more loops were observed ( P < 0.002). In the Caucasian SLE group, seven significant parameters were observed (Figure 3). In the left hand of the SLE group, both the ulnar displacement of the left axial t-triradius ( P < 0.0026) and the composite mea-
DERMATOGLYPHIC STUDY OF SLE
85
pattern in the left hypothenar area, and vertical displacement of the right axial t-triradius (Table 2). The misclassification rate was 26% for the 54 in this group ( P < 0.0012).
DISCUSSION
LEFT
RIGHT
t-TRIR ADllJS SPLACED RADIALLY
\WRIST
The three previous investigations of dermatoglyphics and SLE showed different significant features. Dubes observed a displacement of the axial t-triradius (5). Fraga and Qazi found changes in the pattern frequencies of the fingers (6, 7). Definite racial differences have previously been shown in dermatoglyphic patterns (10, 11). Therefore, because Fraga observed Mexicans, Dubes observed Caucasians, and Qazi observed blacks, their data should be different. The authors of this paper analyzed both Mexican-Americans and Caucasians, and the results in the Caucasian SLE group agree with those of Dubes. He observed an ulnar displacement of the right axial t-triradius, and the present data concur ( P < 0.05). Dubes also found an ulnar displacement of the same measurement predominantly
CREASE
Fig 2. Signijicant parameters in Mexican-A merican groups.
surement (radial displacement/ulnar displacement, P < 0.01 1 ) of the left axial t-triradius were significant, indicating an ulnar shift of the triradius in this group. I n this same hand the SLE group showed a decrease in the log of the a-b ridge count ( P < 0.05) and an increase in hypothenar patterns ( P < 0.036). In the right hand of the SLE group, the authors observed a decrease in the atd angle ( P < 0.042), a decrease in the log of the a-b distance ( P < 0.05), and a proximal displacement of the axial t-triradius ( P < 0.05). The observed number of habitual aborters with SLE was insufficient for analysis. Discriminant analysis was performed using data for each race. In the Mexican-American SLE group the radial measurement of the right axial t-triradius and the presence of a pattern in the left hypothenar area were the only significant parameters in the discriminant function ( P < 0.0028) (Table I ) . The misclassification rate for the 54 in this group was 28%. There were insufficient data to make these measurements in one case in the normal group. In the Caucasian SLE group three parameters were used in the discriminant function: the ulnar measurement of the left axial t-triradius, presence of a
RlST CREASE
LEFT
RIGHT RIDGE COUNT LESS DISPLACEMENT LE
SLE atd
ANGLE LES
DISPLACED ULNAD
PROXIMALLY DISPLACED’ Fig 3. Significant parameters in Caucasian groups.
DUBOIS ET AL
86
Table 1. Discriminant Analysis of Mexican-Americans Using Two Param ters 1 ) Radial displacement of right axial t-triradius 2) Presence of a pattern in the left hypothenar area ~
Dermatographically Analyzed as: ~~
Diagnosed Normal
Diagnosed SLE
Totdl
20 1
n
28
19
26
21
54
~~~
Normal SLE TOTAL
P
21 * < 0.0028
Misclassification rate: 28%.
* Insufficient data in one case.
in the right hand. In the present series the ulnar shift of this same measurement was statistically more significant in the left hand. However several of Dubes’ other less statistically significant findings were not confirmed in the present study. The discrepancy could be due to sampling error, because his Caucasian SLE group consisted of only 7 females. This study did not deal with black SLE patients. No correlation can be drawn. In the Mexican SLE patients Fragaetal observed a diminution of radial loops and digital arches ( 6 ) . In t h e present series a significant increase in whorls (P < 0.002) was found only in the left-hand first finger. Fraga also observed more patterns i n the thenar first interdigital region and the axial t-triradius in t’. The data of the present authors do not substantiate either of these findings. Possibly, Fraga’s Mexican population differed from the Mexican-American group in the Southern California area. Dermatoglyphic differences are known to exist between races (10, 11). In this investigation two features were observed to be statistically significant in separating the two normal groups: the ridge density on the left thumb ( P < 0.046) and the a-b ridge density on the left
hand (P < 0.003). The same approximate result was expected between the Mexican-American SLE group and the Caucasian SLE group. In fact nine significant parameters differentiated the two SLE groups, including the two stated above. These other features included: atd angle right hand (P < 0.014), radial ( P < 0.0033) and ulnar (P < 0.00091) measur,ement of the right axial ttriradius, ulnar measurement of the left axial t-triradius ( P < 0.0022), pattern types on the right fingers (P < 0.05),and the axial t-triradius lateral displacement (composite measurement) in both the right (P< 0.00021) and left (P < 0.013) hands. The discriminant analysis showed a misclassification rate of 25-30%. Although dermatoglyphic changes were associated with SLE, the measurements used were not sufficiently different to suggest their use in the earlier detection of SLE patients. Several genetically inherited diseases show abnormalities in dermatoglyphics. Differences have been observed in the palm prints of patients with Down’s syndrome and those with Klinefelter’s syndrome when they were compared with matched controls (12). Furthermore finger and palm prints have been shown to be genetically controlled (2). Viral infections such as congenital rubella may also modify dermatoglyphic patterns (1 3, 14). Therefore the finding of statistically significant differences between the SLE and normal groups suggests either a genetic abnormality or alteration by a congenitally transmitted virus or perhaps both as a cause of these changes.
REFERENCES 1 . Cummins H, Midlo C: Palmar and plantar epidermal
2.
3. Table 2 . Discriminant Analysis of Caucasians Using Three Parameters I ) Ulnar displacement of left axial t-triradiur 2 ) Presence of a pattern in the left hypothenar area 3 ) Vertical displacement of right axial t-triradius Dermatographically Analyzed as:
Diagnosed Normal
Diagnosed SLE
Total
Normal SLE
19
7 21
26 28
28
54
1
TOTAL
P
