Digital arterial responsiveness to cold in healthy men, vibration white finger and primary Raynaud's phenomenon.

M. Digital arterial responsiveness to cold in healthy men, vibration white finger and pri mary Raynaud's phenomenon. Scand J Work Environ Health 1993;19:271-6. Finger systolic pres sure (FSP) was measured during a standardized cold test in 291 healthy male subjects divided into five age groups. The age groups showed no difference in the change in FSP (FSP%) after cold provo cation at 15 and 100e. In the entire population, the lower normal limits of FSP% were estimated as 76% at 15°C and 63% at 10°C. When a discriminating threshold of FSP% lQo< 60% was applied to the results of the cold test of 31 referents, 65 chain saw workers with or without vibration-induced white finger (VWF), and 20 male patients with primary Raynaud's phenomenon (PRP), the sensitivity of the test to detect digital vasospasm was 84% for VWF and 95% for PRP. In the chain-saw worker group, the positive and negative predictive values of the cold test were 94%. Therefore the measure mentof FSP during cold provocation can be considered a useful laboratory test to confirm Raynaud's symptoms objectively in both groups and individuals.

Several laboratory tests have been proposed to detect digital vasospa sm in subjects affected with primary or secondary Raynaud's ph enomenon. Most of these tests are based on cold pro vocation and the measurement of some indic es of circulatory fun ction before and after local cooling of fingers and hand s with thermometric, laser-Doppler, or plethysmograp hic method s ( 1-3). In occupational medi cine, the measurement of finger sys tolic pressure (FSP) is con sidered a useful vascular test to reveal di gital arterial hyperrespon sivene ss to co ld in workers exposed to chemical and phy sical agent s kno wn to cause vasospastic syndromes, such as hand-arm vibration , extreme cold environment, vinyl chloride, and arsenic (4)(5)(6). However, owing to the wide variability of the response of the digital arteries to cold and to the shortage of referenc e dat a from normal individuals, different estimates of the acc uracy of FSP mea surement in the obj ecti ve diagnosis of Raynand' s phenomenon have been reported by vario us authors (I , 3, 4, 7, 8).
In this study FSP measurement during local cooling was perform ed in a large samp le of health y male subj ects of working age to investigate the phy siological reaction to cold in the digital arteries of normal individuals. Standardized laboratory method s for FSP measurement and cold pro vocation were used to obtain normative values for cold-induced chang- es in FSP in healthy subj ects. In addition, the se nsitivity of the cold test to detect digital vasospasm was assessed in a group of men affected with primary Raynaud ' s phenomenon and in a popul ation of chainsaw work ers at risk for vibr ation-induced whit e finger. For this latter group, the posit ive and negati ve pred ictive valu es of the cold test wer e also evaluated.

Subjects
Two hundred and ninety-one healthy men , aged 20 to 69 years, underwent a medi cal interview and a complete clin ical examination at the vascul ar laboratory of the Institute of Occupational Health , Trieste. None of the subj ects had any cardiovascular or neurolo gical abnormalities in a physical examination , and none of them we re on any form of med icat ion. All of them had a negati ve fam ily history of cold hypersensitivity in the hands or of constitutional white fingers.
The study population was divided into the following five groups according to age : 20-29 years (N= 44) ,30-39 yea rs (N = 64), 40--49 yea rs (N = 76), 50-59 years (N = 72) , and 60-69 years (N = 35). Smoking and drinking habits, expressed in term s of grams of tobacco and alcohol per day , were not different amon g the age groups . White-collar workers (clerks, officers) represented about one-third of the sample, while the remaining subjects were bluecollar workers (construc tion work ers , mechanics, electricians) not expo sed to hand -arm vibration or other agents causi ng Raynaud' s phenomenon of oc-cupational origin. A group of 65 forestry workers using chain saws (mean age 44.7 years) and a reference group of 31 manual workers not expo sed to vibration (mean age 44.0 years) were also examined. The duration of exposure to chain-saw vibration averaged 11.3 years in the fore stry worker group . Digital vaso spastic symptoms [ic, vibration-induced white finger (VWF)] were rated according to the Stockholm Workshop scale (9) . Finally, 20 male patients diagnosed consecutively as having primary Raynaud's phenomenon (mean age 37.1 years) were included in the study . The diagnosis of primary Raynaud ' s phenomenon was made according to the clinical criteria suggested by Allen & Brown (10). The patients with prim ary Raynaud 's phenomenon had bilateral involvement of most fingers, and 40 % of them had attacks of both fingers and toes.

Fing er systolic pressure and cold provocation
The cold test was performed with the subj ect in a supine position after a rest of 30 min in a thermostated room with an ambient temperature of 22-23°C. The subjects wore light clothing durin g the test. Digital systolic blood pre ssure was mea sured after cold provocation of a test finger according to the cooling technique proposed by Nielsen & Lassen (II). A double-inlet plastic cuff for both air filling and watcr perfusion was plac ed around the middle phalanx of the third left finger of the healthy referents and the cha in-saw workers without VWF. In the subjects with VWF or primary Raynaud's phenomenon, the most affected digit was cooled. The test finger was thermo stated with water circulating at 30, 15, or lOoC by using a dig it cooling system (Medimatic AlS, Copenhagen, Denmark). Two air-filled cuffs were applied on the proximal phalanx of the test finger (for ischemia during cooling) and on the middl e phalanx of a reference finger (usually the fourth finger). The cold test was performed by pressurizing the air cuffs at a suprasystolic level (2 10 mm Hg or ",28.0 kPa) and perfusing the water cuff initially at 30°C and then at 15 and lOoC. Afte r 5 min of ischemic cooling, the digital systolic blood pressure was measured on the test and reference fingers by a plethysmographic technique. Arm systolic pressure was measured by a mercury sphygmomanometer using a standard cuff. For each subjec t, the foll owing two pre ssure indices were deriv ed from the arm and finger systolic pres sures: In order to avoid nicotine-induced vasoconstrictive effects on the digital arteries, tobacco users had refrained from smoking for at least 4 h before the testing.

Statistical analysis
The FSP measurements were expressed as means and standard deviations or as medians and ranges. The lower normal limits of the FSP parameters were calculated as the mean -(SD . t o . o ) ' where SD is the standard devi ation and t is the percentile of the Student's t distribution. The differences among the mean values were tested by an analysis of variance and covaria nce. Two or k independent samples of nonnormally distributed data wer e compared by the Mann-Whitney test and the Kruskal-Wallis test, respectively. In both the parametric and nonparametric one-wa y analysis of variance , mult iple co mpari-SOn tests were used to compare pairs of groups. A P-value of 0.05 (two-sided) was chosen as the limit of stati stical significance. The sensitivity, specificity, and predictive value of the cold test to diagnose Raynaud's phenomenon obje cti vely were evaluated by the receiver operating characteristics (ROC) analysis. Table I shows that among the healthy men the FSP in the test and reference fingers after warming to 30°C was lower in the younger subjects (20-29 years) than in the older ones (50-69 years) (P < 0.05). However, when the FSP was normalized to the arm systolic pressure, the digital pres sure index (DPI 3 0") was similar in all of the groups. Within each age group , no difference was observed between the FSP measured on the test and reference fingers. No relation was found between FSP 30 , of the test finger and smoking and drinkin g habits . Finger cooling provoked a significant decrease in FSP in the test finger. In the whole group, the reduction in FSP averaged 2.3 mm Hg ("' 0.31 kPa ) at 15°C and 7.6 mm Hg ("'1.01 kPa) at ro-c (P<0.005). Table 2 presents the mean values, standard deviations, and lower normal limit s for FSP %, DPI I 5 . , and DPI !O". The analysis of varia nce showed no difference in the pressur e param eters among the various age groups, except for DPI IO • (P<0.05). In the entire group, the low er normal limit s for FSP%15' and FSP %IQ' were 76 and 63%, respecti vely. After control for age, the FSP parameters at lOoC were found to be lower for the smokers than for the non smokers (table 3), but the difference was significant only for DPI ,oo (P < O.05). Among the smokers , a significant inverse relationship was observed between the decrease in FSP% and DPI IQ' and daily tobacco consumption (P < O.Ol) . Table 4 shows the results of the cold test for the 31 manual workers in the reference group, the 65 chainsaw workers with or without VWF, and the 20 pa-Scand J Work Envi ron Health 1993, vol 19, no 4 tients affected with primary Raynaud' s phenomenon. Among the chain-saw operators, the prevalence of VWF was 29.2%. Of the 19 forestry workers suffering from VWF, seven ( 10.7%) were in stage 1, seven (10.7%) were in stage 2, and five (7.7%) were in stage 3 according to the classification of the Stockholm Workshop (9). The latency inter val for VWF (ie, the length of exposure to vibration before the first episode of finger blanching) averaged 9.4 (SD 6.8) years. Finger cooling to 15 and lOoC caused a sig- Table 1. Arm and finger blood pressure in the healthy male subjects during the provocation test with water at 30·C. (ASP = arm systolic pressure, FSP t = finger systolic pressure in the test finger, FSP,ef = finger systolic pressure in the reference fi nger, OPI = dig ital pressure index, calculated as the rat io of FSP, to ASP) • P <0.01 (one-way analys is of variance fo r the difference among age groups).   nificant decrease in FSP% and DPI among the chainsaw workers with VWF and the patients with primary Raynaud ' s phenomenon in compariso n with the subjects with no vasospastic symptoms. The patien ts with primary Raynaud' s phenomenon exhibited more powerful digital vasospasm at lODe than the professional chain sawyers with VWF (P <0.002) . A multipl e compariso n test showed no differences in the change of FSP and DPI between the referents and the chain-saw workers without VWF. Table 5 shows that the VWF subj ects with more severe symptoms (stage 3) experie nced a stronger peripheral vasoco nstric tion at 10°C than those with mild or moderate symptoms (stag e I and 2). It is wort h noting that age was similar in the two VWF subgroups, while the two subgroups differed significantl y with respect to the durati on of exposure to vibration. No difference in FSP% and DPI]O" was observed between the patien ts with primary Raynaud' s phenomenon and the chain sawyers with stage 3 of VWF . Tot al clos ure of the digital arteries during local cooling to lODe (ie, zero blood pressure in the test finger) occurred in five VWF workers (26.3 %) and 15 primary Raynaud ' s phenomenon patients (75.0 %). Among the VWF chain sawyers showing the clos ing phenomenon at lODe, four (80%) were in stage 3. One chain-saw worker with peripheral sensorineur al distu rbance s alone had zero pressure 274 in the finger at both 15 and lODe, a finding indicating a preclinical condition of Raynaud ' s phenom enon. When age and tobacco consumption were allowed for, FSP% I0 0 was found to be inversely related to the duration of exposure to vibration among the chain sawyers (P =0.003).

Sensitivity, specificity and predictive value of the cold test
The sensitivity and specificity of the cold test to detect digital vasospasm in Raynaud ' s phen omenon was assessed with the use of FSP%\0"<60% as the discrimin ating threshold between normal and pathological responses of the digital arteries to cold provocation (table 6). Abnormal cold reactions in the digital vessel s were observed in none of the referents (specificity among the unexposed referents bein g 100%) and in one chain sawyer without VWF (specificity among the "exposed" referents being 97.8%). Three workers affected with VWF and one patient suffering from primary Raynaud' s phenomen on showed fal sely negative results (ie, normoreactivi ty to cold with Raynaud' s symptoms), a finding indicating that the sensitivity of the cold test was 84.2 and 95.0%, respectively. In the diagnosis of VWF, the predictive value of a positive test (FSP%,0"< 60%) was found to be 94.1 %, while the predictive value of a negati ve test was 93.7 %.

Discussion
The results of this study indicate that the change in FSP during local cooling to 15 and 10°C was similar in normal male subjects aged 20 to 69 years . Therefore, age does not see m to influence the reaction of digita l arteries to cold in healthy men. OPI,O" was found to be lower in the age range 60--69 years than in the younger age groups. This finding may be ascribed to factors related to both the sample size and the greater increase in arm systolic pressure at IDoC observed in the older subjects. Cold-induced vasoconstriction was stronger in the smokers than in the nonsmokers. It has been reported that among VWF workers tobacco users show more severe vasospastic symptoms and an increase d hyperreactivity to cold provocation than nonusers (12). These findings suggest that smoking has a harmful effect on digital circulation in both normal subjec ts and VWF workers.
In this study, the results of the cold test indicated that the variance of FSP % at 15 and I DoC was smaller than that of the OPI. Since the change in FSP in the cooled finger was corrected for the change in systemic arterial blood pressure, the FSP% can be considered a more stable pressure parameter than OP1, which closely depend s on the variations in arterial systolic pressure during cold provocation. In a previous study of the repeatabil ity of finger systolic pressure measurements ( 13), it was found that for five normal men tested for five consecutive days the coefficient of variation for repeated determi nations of FSP % at 15 and 10°C averaged 6.4 (SO 2.2)%, a figure lower than the mean value of 7.3 (SO 3.5)% observed for the corres ponding DPI. These findings, as well as the resul ts of other studies (4,14), suggest that FSP % is a robust and repeatable index of digital circ ulatory function during cold provocation.
In this investigation, FSP% at 15 and 10°C could differentiate between patients with VWF and primary Raynaud ' s phenom enon and other subj ects not affected with vasospastic symptoms. Furthermore, the cold test could separate chain-saw workers with mild and moderate forms of VWF (stage 1 and 2) from those with more severe finger blanchin g attacks (stage 3). In primary Raynaud ' s phenomenon the cold response of digital arte ries was more exaggerated than in VWF. A similar finding has been reported in severa l studies of Raynaud ' s phenomenon (8,15,16). Experimental investigations have pointed out a hyperactivity of the centra l sympathetic nervous system during cold exposure and postural stress in both primary Raynaud' s phenomen on and VWF patients ( 17,18). In a study of the sympathetic vasoco nstrictor reflex mechanism in Raynaud ' s phenomenon (16), primary Raynaud ' s phenomenon subjec ts showed an increased vasoconstriction to vibration when compared with VWF workers. It has been suggested that prolong ed use of vibrat ing tools can induce either an adaptation of the central nervous system to vibration or a periph eral sensory neuropathy which attenu ates, at least partially, the vasoconstri ctor response of the digital arteries in VWF workers ( 15,16,19). However, the result s of the present study indic ate that the augmented vascular reactivity to cold in the primary Raynaud' s phenomenon group, compared with the VWF group, may also have been due to differences in the clinical stage of the vasospastic syndrome, as cold-induced digital arterial hyperresponsiveness was not found to be significantly different between the patients with primary Raynaud ' s phenomenon and the vibration-exposed workers with severe VWF symptoms. It has been reported that vibration-exposed workers may exhibit an abnormal digital arterial cold response before the appearance of VWF symptoms (4) . In this investigation, the closing phenomenon of the digital arteries was observed in one chain-saw worker with a negative history of finger blanch ing attacks. Th is finding confirms that FSP measurement durin g local cooling is useful in disclosing peripher al vascular hyperreactivity in vibration-exposed workers without subjective Raynaud ' s symptoms.
In this study, a lower normal limit of 75% for FSP% failed to discriminate on an individual ba-I~h' h sis in the vibration-e xposed group because of a rg proportion of falsely negative result s among the chain sawyers with VWF (84%). On the contrary, the sensitivity of the cold test at 15°C was satisfactory for the primary Raynaud ' s phenomenon group (86.3%). It is likely that local cooling to 15°C is not a sufficient cold stimulus to provoke digit al vasospasm in VWF workers, mainly among those with mild or moderate vasospastic symptoms. The analysis of the ROC curve for FSP%I00showed that a value of 60% was the optimal threshold with which to 275 Scand J Work Environ Health 1993, vol 19, no 4 obtain a high true positive fraction (sensitivity) and a negligible false positive fraction ( l -speciticity) in both the primary Raynaud's phenomenon and VWF groups. It is worth noting that the discriminating threshold of FSP% lOo<60% is in close agreement with the lo wer limit found for the normal population examined in thi s study (63%). Moreover, th e proposed threshold is al so consistent with those suggested by other authors who reported lower normal limits o f 59-66% at lOoe in smaller sa m p le s of male referents (5,7,20).
Olsen et al (4) have suggested the ri gid criterion of zero pressure (closing phenomenon of th e digi tal arteries) to d iagnose Raynaud' s sy ndro me . They consi dered th at both local and body cooling are needed to provoke a com plete digital vaso spasm in Raynaud patients. In this study body c ooli ng was not pe rformed o wing the lack of consent by most referen ts a nd vibration-exposed workers who felt the procedure was uncomfortable. However , the findings of thi s investigation and clinical experience suggest that the diagnost ic crite rion adopted (FSP% lO o < 60%) is sufficiently stri ct to detect ab nor m al cold re sponse in the di git al arteries of mo st subj ects with a true hi story of prim ary R aynaud' s phenomenon or VWF. This opinion seems to be supported by the results reported in table 6 regarding the efficiency of the cold test to diagnose Raynaud' s phenomenon objecti vely in a g ro up of chain sawyers w ith a known VWF pre valence. The di scrim in ating th re shold of FSP% \0"<60% appears to be an ap pro priate diagnostic criterio n to re veal VWF, as in th e chain-saw group the predictive value o f a po siti ve te st -denoting how often the cold test was correct when its result was positive -was as high as 94%. A similar fi gure was ob served for the predicti ve value o f a ne gativ e te st (i e, correctness of neg ati ve re sults).
In conclusion, th e finding s o f this study suggest that the change in FSP during standardized cold provo cati on at 10°C is a useful laboratory te st to di fferentiate between he althy subjects and pat ients wi th primary Raynaud's phenomenon or VWF both on a group ba sis and on an individual basis. Th erefore the pre ssure parameter FSP%!O" can b e us ed not only in epi d e m io logic su rv eys, but als o for medicoleg al problems and in suranc e co mpe nsation purposes.