Objective. To examine the clinical presentation and disease associations of Raynaud’s phenomenon (RP) in children and adolescents.
Methods. A systematic retrospective chart review was conducted of 123 cases drawn from 2 computerized databases at the Children’s Hospital of Boston. Participants aged <19 years with episodic reversible color changes in the extremities were examined. Case records were analyzed for clinical presentation, disease associations, and physical examination and laboratory findings.
Results. In contrast to the findings of smaller pediatric series reported to date, the large majority of our patients—approximately 70%—did not have a recognized underlying connective tissue disease. For both primary and secondary RP, approximately 80% of patients were female, and mean age of onset was similar in the 2 groups. Biphasic or triphasic color changes were less common than monophasic changes and were no more common in secondary than primary RP. Findings predictive of secondary RP were limited to the presence of antinuclear antibodies and abnormal nailfold capillaries. Antiphospholipid antibodies were found at some time in at least 21% of patients with both primary and secondary RP.
Conclusions. RP in children, as in adults, principally affects girls and is frequently free of association with connective tissue disease. Antinuclear antibody positivity and abnormal nailfold capillaries correlate with secondary disease. Antiphospholipid antibodies are surprisingly common, a new finding of uncertain implications.
- Raynaud’s phenomenon
- antinuclear antibodies
- nailfold capillaries
- antiphospholipid antibodies
Color changes of the distal extremities in response to cold or stress were first described by Raynaud in 1862.1 Raynaud’s phenomenon (RP) may occur in the absence of accompanying disease, termed primary RP, or may be secondary to an underlying disease such as scleroderma or lupus.2 RP is common in the adult population, with estimates of prevalence ranging between 1.8% and 21.1% in women and 0.5% to 16% in men.3 More than 80% of these individuals have primary RP.4
Data on RP in children are limited. Although children and young adults have been included in large observational series,5,6 detailed reports number <100.7–16 The largest series comprised 27 patients seen for RP in a pediatric rheumatology clinic in Toronto, two thirds of whom had either definite or probable connective tissue disease at presentation.16 In line with this literature, textbooks describe RP in children as a relatively rare disorder that is usually secondary to an underlying disease.17
The interpretation of reports of RP is complicated by problems of definition. Maurice Raynaud described the “classic” triphasic color change of pallor, cyanosis, and erythema limited to the digits. However, only a fraction of patients with RP have this pattern, even among Raynaud’s original 25 patients.1,5,6 Our clinical impression has been that RP-like phenomena in children are highly heterogeneous, relatively common, and clinically more benign than has been suggested in the literature. We elected to investigate this impression by reviewing patients seen at the Children’s Hospital of Boston with reversible color changes in the extremities.
We identified patients using 2 computerized databases. The hospital discharge database was surveyed from July 1, 1991, to July 1, 2001, for the following International Classification of Diseases, Ninth Revision codes: Raynaud’s (443.0), acrocyanosis (443.89), vasculitis (447.6), vascular disease (459.9), systemic lupus erythematosus (SLE; 710.0), scleroderma (710.1), Sjogren’s syndrome (710.2), and mixed connective tissue disease (710.8). The rheumatology outpatient clinic database, cataloging principal medical diagnoses and therapies as abstracted by secretarial staff from outpatient clinic notes, was surveyed from its inception on December 1, 1998, to July 1, 2001, for the following diagnoses (no International Classification of Diseases, Ninth Revision classification used): Raynaud’s, acrocyanosis, vasculitis, SLE, scleroderma, Sjogren’s syndrome, mixed connective tissue disease, antinuclear antibody (ANA), antiphospholipid (APL) antibody, nifedipine therapy, and prazosin therapy. The medical records of the approximately 600 patients identified by database search were then reviewed for the presence of RP-like phenomena. In an attempt to avoid skewing our findings as to the clinical presentation of RP, inclusion criteria for data collection and analysis were kept deliberately broad: 1) history of episodic reversible color changes in the extremities and 2) age <19 years at onset of color changes.
Clinical and laboratory data were extracted from computerized records, paper charts, and the hospital laboratory computer system in a standardized manner. Information obtained included demographic information, age at onset of symptoms, age at first diagnosis at Children’s Hospital of Boston, and age at last informative encounter within the Children’s Hospital system (defined as an encounter at which the presence of a serious diagnosis was likely to be recorded, eg, an emergency department or outpatient clinic visit but not a note recording a prescription refill). At each of these time points, the treating physician’s diagnosis relevant to the color changes was recorded, including connective tissue diseases, other conditions, and the diagnosis of acrocyanosis. Clinical information regarding medical history included associated medical illnesses, history of chemotherapy, use of oral sympathomimetics or β-blockers, and history of frostbite. Details of the color changes were noted, including color order, pattern of finger involvement, extent of changes up the extremities, involvement of other body parts, triggers, frequency, adverse consequences (including pain and tissue damage), and treatment. Physical findings and associated laboratory abnormalities were recorded, including nailfold capillary changes, digital tissue loss, livedo reticularis, ANA, specific autoantibodies, and APL antibodies. Missing data were coded as such in the database.
For the purposes of this study, patients who met entry criteria were divided into 1 of 3 mutually exclusive and collectively exhaustive diagnostic categories. Primary RP was defined as episodic reversible color changes in the extremities without established or suspected connective tissue disease. Secondary RP was defined as episodic reversible color changes in the extremities in the setting of a connective tissue disease known to be associated with RP: SLE, systemic sclerosis, mixed connective tissue disease, overlap connective tissue disease, and suspected early connective tissue disease. Because APL antibodies are not an established cause of RP, patients with RP and APL antibodies in the absence of established connective tissue disease were classified as having primary RP, even when the treating physician suspected a causal relationship. Other diagnosis was defined as episodic reversible color changes in the extremities as a result of a well-defined cause not included in this list of connective tissue disorders (eg, vasculitis). Data were analyzed using SPSS (SPSS, Inc, Chicago IL). The Children’s Hospital Institutional Review Board approved the chart review.
Our population consisted of 123 children, 80% of whom were female. At last follow-up, 69% carried the diagnosis of primary RP, 28% carried the diagnosis of secondary RP, and 3% carried other diagnoses (1 patient each with polyarteritis nodosa, thromboembolic disease, vascular compression from a Baker’s cyst, and fibromuscular dysplasia). Mean age at symptom onset was 12.3 ± 4.3 years, and mean age at diagnosis was 13.4 ± 4.0 years. Age at onset and age at diagnosis were not significantly different between primary and secondary cases. Length of follow-up was significantly longer in patients with secondary RP (P = .036; Table 1). A plot of age distribution at onset is shown in Fig 1.
Characterization of Color Changes in the Extremities
The pattern of color changes was recorded by the examining physician in 93% of cases (n = 114). Number of color phases did not differentiate primary from secondary RP (Table 2). Among these 114 patients, secondary RP was present in 27% of patients with monophasic changes, 22% with biphasic changes, and 20% with triphasic changes. Monophasic color changes were white, blue, or a mottling of colors including white or blue (no patient identified by our search had isolated erythema). Biphasic color changes in primary and secondary RP were essentially equally divided between white→red and blue→red, whereas triphasic color changes were equally divided between blue→white→red and white→blue→red (data not shown). In approximately 5% of both primary and secondary RP, color changes spared the fingers and were limited to the lower extremities. Toes were noted to be affected in 49 patients (58%) with primary RP and 13 patients (38%) with secondary RP, although this may underestimate the true prevalence because it was not possible to determine how often such involvement was sought by the examining physician. Color changes that extended beyond the wrists or ankles was reported by 10 patients (8%); all had primary RP. The frequency of episodes was recorded too inconsistently to be meaningful. Self-reported triggers for color changes in primary and secondary RP are listed in Table 3; multiple triggers were common. Boys were similar to girls in clinical presentation (data not shown).
Morbidity and Treatment
The color changes were noted to be experienced as painful (broadly defined to include tingling, numbness, and other disagreeable sensations) in 47 (55%) cases of primary RP and 10 (29%) cases of secondary RP. Reanalyzed to exclude cases in which no comments on sensation were recorded by the examining physician, pain was present in 67% of primary RP and 63% of secondary RP. Adverse consequences of RP, aside from discomfort, were experienced by 7% of patients with primary RP and 18% of patients with secondary RP (χ2 = 2.8, 0.10 > P > .05; Table 4). One patient died of non-RP complications of mixed connective tissue disease (scleroderma predominant) after a difficult course including digital amputations and intractable digital pain. Treatment for RP, beyond routine education in environmental control, was administered to 51% of patients with primary RP and 44% of patients with secondary RP. The most frequent therapies included calcium channel blockers, aspirin (used in patients with APL antibodies), and biofeedback. Overall, boys received therapy more frequently than girls (72% vs 45%, χ2 = 5.9, P < .025).
Clinical and Laboratory Findings
Abnormal nailfold capillaries and the presence of ANA have been noted to associate strongly with secondary RP in adults.18 In our population of patients for whom information was available, nailfold capillaries were borderline or abnormal in 23% of primary RP and 68% of secondary RP (χ2 = 13.2, P < .001; Table 5). ANA was positive in 25% of primary RP (n = 19 positive of 76 tested) and 85% of secondary RP (n = 29 positive, all 34 tested; χ2 = 34.7, P < .001). Boys were similar to girls in these characteristics (data not shown). The distribution of ANA titers is shown in Fig 2. Of patients in whom a precipitin panel was checked—including where measured double-stranded DNA, Smith, Ro, La, RNP, and SCL-70 and assumed to be negative when ANA was negative—this panel was positive in 6% of primary RP (4 patients of 69 tested) and 67% of secondary RP (20 patients of 30 tested; χ2 = 41, P < .001). Digital pits were found in 2 (2%) patients with primary RP and 5 (15%) patients with secondary RP. Livedo reticularis was noted in 14 (16%) patients with primary RP and 3 (9%) patients with secondary RP. APL antibodies were equally common in patients with and without livedo (data not shown).
APL antibodies have been associated with RP and related disorders in adults.19 Among patients in whom 1 or more APL antibodies were measured, 36% of patients with primary RP were positive at least once, compared with 30% with secondary RP (6 patients with SLE, 1 with mixed connective tissue disease). If all patients who were not tested are assumed to have been negative, then the proportion of patients who were positive for APL antibodies at some time is still substantial at 21% in both primary and secondary RP. The specific antibodies noted are shown in Table 6.
The 18 patients with primary RP and positive APL titers were investigated more closely for evidence that they constitute a separate etiologic subgroup. They did not differ in age of onset from the 32 patients with primary RP and negative APL antibodies (data not shown). They tended more often to be female (89% vs 66%, χ2 = 3.2, 0.10 > P > .05) and to have consequences of their RP, such as ulcers or need for hospitalization (17% vs 3% for adverse consequences as listed in Table 4, χ2 = 2.9, 0.10 > P > .05), and they were more likely to experience pain or discomfort from their RP (82% vs 62%, χ2 = 4.1, P < .05). Most of these patients were treated with low-dose aspirin. In 3 cases, the APL antibody was of questionable importance (anticardiolipin immunoglobulin [Ig] M titer 10–15 IgM phospholipid units [MPL]), whereas in 9 of 18 cases, the antibody titers dropped to undetectable levels over the course of follow-up. Two individuals had acute digital ischemia in the setting of high-titer APL antibodies. The first was a 7-year-old girl with acute onset of episodic reversible monophasic color changes in 3 toes of her right foot, anticardiolipin IgG >248 IgG phospholipid units, IgA 33.5 APL, IgM 29 MPL, and a positive lupus anticoagulant. ANA was negative. She did well on aspirin but relapsed when her aspirin was discontinued, although her APL antibodies became undetectable with time. Four years later, she is generally healthy but requires aspirin in cold weather. The second was a 17-year-old boy with episodic cyanosis of the left fifth finger and, several months later, 2 fingers on the right hand requiring intensive care unit admission for intravenous prostaglandins. ANA was positive at 1:2560, and dsDNA was positive, but there was no other clinical evidence for SLE. His anticardiolipin IgG was >70 IgG phospholipid units, and IgM was 24 MPL, and he tested positive for mycoplasma, with waning of his APL titers over time. He returned to his country of origin and was lost to follow-up.
Nine patients classified as primary RP for the purposes of this study had received the diagnosis of “acrocyanosis” by the attending pediatric rheumatologist. These 7 girls and 2 boys were investigated further for characteristics that distinguish them from other patients with RP. Four were age 2 or younger at onset of symptoms, and 5 were between 11 and 16 years of age. These 4 young patients account for the apparent “bimodality” of the age of onset of primary RP in Fig 1. The color changes noted by these patients were uniformly symmetric; were mono- or biphasic; and tended to involve the whole hand, foot, or both under the influence of cold or without apparent trigger. Extension beyond the wrist was reported by 3 of 9 patients, and extension above the ankles was noted in 6 of 9 patients. Discomfort was reported by 1 patient, and the only recorded adverse consequence was an episode of subjectively slow wound healing.
Diagnoses associated with primary RP, as attributed by the examining physicians, included weight loss or being underweight (n = 4), history of frostbite (n = 1), and use of sympathomimetics for attention-deficit/hyperactivity disorder (n = 10). Although physicians recommended discontinuation of sympathomimetics in 1 patient, no patient showed a clear correlation between symptoms and use of medication. Two patients had recurrent vasovagal episodes, and 1 had autonomic instability associated with ornithine transcarbamoylase deficiency. One patient with secondary RP also had celiac sprue. No patient was on β-blockers, and no patient had a history of receiving chemotherapy for cancer.
Classically, patients with RP describe triphasic color changes in the digits, with blanching (white) giving way to cyanosis (blue) followed by reactive hyperemia (red).20 However, not every patient has all 3 phases. The UK Scleroderma Study Group defines “definite” RP as 2 color changes in response to cold, whereas uniphasic color change accompanied by numbness or paresthesia is termed “possible” RP.21 Other authorities stress sharp demarcation of borders as the sine qua non of RP, regardless of whether multiple color phases are present.2 Of 2 recent large-scale trials for therapy of RP, 1 required patients to have triphasic color changes (order not specified)22 and the other accepted any patient with episodic digital blanching or cyanosis in response to cold or emotional stress.23
Two other entities fall within the spectrum of episodic digital color changes: acrocyanosis and chilblains (pernio). Well recognized in the newborn infant, acrocyanosis may occur at any age as a usually painless, persistent, clammy cyanosis of the distal extremities, typically with poorly defined borders.24,25 Acrocyanosis typically but not invariably waxes and wanes with cold and emotional stress and is thought to represent an exaggerated vasomotor response.17 The prognosis is benign. Chilblains refers to episodic color changes after exposure to severe cold, typically with nodules, and is thought to represent spasm-induced vessel and tissue damage.26 Endothelial damage in lupus can cause a similar phenomenon in the absence of cold, called lupus pernio or chilblain lupus.27 Because frostbite and SLE can also result in classic RP,28 it is clear that these conditions may overlap.
Pathophysiologically, reversible color changes in the extremities can result from vasospasm, fixed partial obstruction, altered flow characteristics of the blood, or a combination of these mechanisms. Peripheral vasospasm is a physiologic response to cold and occurs in all normal individuals. Excessive vasospasm can result from more or less subtle dysregulation among the determinants of vascular tone: autonomic stimuli, circulating catecholamines, endothelial-derived factors, and the response characteristics of the vascular smooth muscle.2 Injury to endothelium or other tissue may account for RP observed with chemotherapy,29 frostbite,28 smoking,4 APL antibodies,19 or chronic use of vibrating hand tools.30 Drugs such as β-blockers may contribute to imbalance between vasoconstriction and vasodilation.4
The effect of spasm is magnified when flow is otherwise partially obstructed, as by atherosclerosis, structural anomaly, vasculitis, or thromboembolism. Obstruction is an important mechanism in secondary RP. In a large series of adult patients with RP, two thirds of 279 patients with secondary RP exhibited obstructed finger blood flow as measured by photoplethysmography in a warm room.31 In connective tissue diseases, the mechanism of this obstruction is most frequently one of progressive intimal fibrosis, best documented in scleroderma.2 In this setting, vasospasm of whatever cause can result in abrupt and complete vascular occlusion.
Given the variety of pathophysiologic mechanisms in RP, it is not surprising that the clinical presentation is nonspecific. This series confirms that the number of color phases does not distinguish primary from secondary RP in children. Only a minority of patients, even with secondary RP, had triphasic changes. The extent that color changes extend proximally along the extremity may have limited discriminatory value: no patient who reported color changes proximal to wrist or ankle had anything other than primary RP as defined in this study (most of these patients were termed acrocyanosis by the examining physician).
Our series documents that the large majority of RP in children is primary, as is the case in adults. Predictors of an underlying connective tissue disease identified in this study include a positive ANA (85% vs 25%; P < .001), positive precipitin panel (67% vs 6%; P < .001), and abnormal nailfold capillaries (borderline or definitely abnormal 68% vs 23%; P < .001). Ninety-three percent of all patients with an underlying connective tissue disease had either a positive ANA or some nailfold abnormality, although 39% of patients with primary RP (33 of 85) were also found to be abnormal in one or the other. The difference in the rate of consequences of RP between primary and secondary disease (18% vs 7%) approached but did not reach significance. APL antibody positivity did not distinguish primary from secondary cases.
Over the course of this study, we noted 8 cases of patients who originally were believed to have primary RP and subsequently received a diagnosis of having another disorder. Seven patients developed connective tissue disease (3 with undifferentiated or overlap connective tissue disease, 2 with suspected early connective tissue disease, and 1 each with SLE and mixed connective tissue disease), and 1 was found to have compressive vasculopathy from a Baker’s cyst causing color changes in the leg. The 7 patients with connective tissue disease all were female, and 6 of 7 had either abnormal nailfold capillaries or a positive ANA. Because the bulk of our cases of primary RP came from the newly instituted rheumatology clinic registry and are therefore recent, it may be that more will convert over time. Our follow-up of primary RP cases was generally short (1.3 ± 2.1 years), and Spencer-Green’s adult data documented conversion from primary to secondary RP in 12.6% of patients a mean of 10 years after onset of symptoms.18
APL antibodies have been noted to associate with RP in rheumatoid arthritis32 and are more prevalent in adult patients with secondary RP (although not primary RP) than in controls.19 Case reports have associated APL antibodies with digital ischemia in children.33 Our trial extends these data by documenting that APL antibodies are a surprisingly common finding in children with RP. Of children with RP tested for APL antibodies, 36% with primary RP and 30% with secondary RP had at least 1 abnormal titer (21% in both groups of patients not tested are considered negative). No control group was assessed, but this finding is far in excess of the 1% to 5% reported for the general population34 and for a pediatric control population.35 Ascertainment bias is unlikely to have accounted for these results among patients with primary RP, because negative tests were not generally repeated. Although a causative role cannot be assumed, APL antibodies are known to bind and activate endothelium36 and might therefore result in endothelial damage and/or thrombosis in the digits. Such injury could manifest as RP, even if the antibodies were present only transiently. Alternately, APL antibodies could result from exposure of neoantigens via endovascular injury or infectious insult and therefore represent no more than a marker of disease. In our series, no patient with APL antibodies developed a large-vessel clot or miscarriage, so the pathologic significance of the antibodies and the necessity for therapy remain undefined.
Four children (3 girls and 1 boy, ages 13–18 years) were noted to have primary RP in the setting of weight loss or low body weight requiring medical management (gastroenterology/nutrition referral in 3 patients, hospitalization for eating disorder in 1 patient). One patient with RP secondary to suspected early connective tissue disease had celiac sprue resulting in mild malnutrition. It is interesting to note that anorexia nervosa has a known association with acrocyanosis and RP.37,38 Earlier series also documented a disproportionate percentage of “underweight” individuals among patients with RP: 33% of men and 50% of women.39 Physicians who evaluate adolescents with RP would do well to bear this association in mind.
Our study is limited by its retrospective nature and case-finding methodology. Documentation of symptoms was likely to have been more complete in primary RP than in secondary disease, where other aspects of disease management frequently take priority. As a result, chart review may not reflect the full clinical presentation of secondary RP. We were unable to assess Wigley and Flavahan’s2 suggestion that a clear line demarcating the zone of color change distinguishes true RP, because physicians did not generally record this observation in clinic notes. Because most of our patients were identified through the rheumatology clinic database or by rheumatologic discharge diagnosis, this series likely overestimates the prevalence of underlying connective tissue disease in children with RP. If so, then the proportion of pediatric RP that is primary and hence usually clinically benign is even higher than the approximately 70% documented here. Confirmation of this result awaits population-based, prospective data collection.
This study represents the first attempt to define the clinical presentation of RP in the pediatric age group. Our findings suggest that pediatric RP is similar to adult disease in being most commonly primary, although association with connective tissue disease also can occur. Pattern of color changes and age of onset do not distinguish primary from secondary RP. Rather, as in adults, ANA and nailfold capillary examination are the most useful discriminatory tests. APL antibodies are a surprisingly frequent finding in pediatric RP, whether primary or secondary, although an causative role for these antibodies remains to be defined.
This work was supported in part by the Samara Jan Turkel Center for Pediatric Autoimmune Disease and by National Institutes of Health Grant P60AR36308.
We thank Drs J. N. Katz and M. H. Liang for helpful comments on the manuscript.
- Received March 22, 2002.
- Accepted September 25, 2002.
- Reprint requests to (P.A.N.) Rheumatology, Fegan 6, Children’s Hospital, 300 Longwood Ave, Boston MA 02115. E-mail:
- ↵Raynaud M. On Local Asphyxia and Symmetrical Gangrene of the Extremities. (trans. Thomas Barlow). London, United Kingdom: The New Sydenham Society; 1888
- ↵Voulgari PV, Alamanos Y, Papazisi D, Christou K, Papanikolaou C, Drosos A. Prevalence of Raynaud’s phenomenon in a healthy Greek population. Ann Rheum Dis.2000;59 :206– 210
- ↵Gifford RW, Hines EA. Raynaud’s disease among women and girls. Circulation.1957;14 :1012– 1021
- ↵Gunteroth WG, Morgan BC, Harbinson JA, Mullins GL. Raynaud’s disease in children. Circulation.1967;34 :724– 729
- Sayre JW. Raynaud’s disease presenting in a 5-month-old-male infant. Pediatrics.1973;52 :412– 415
- Emert H, Schaller JG. Raynaud’s phenomenon in childhood. Arthritis Rheum.1977;20 :363
- Burns EC, Dunger DB, Dillon MJ. Raynaud’s disease. Arch Dis Child.1985;60 :537– 541
- ↵Cassidy JT, Petty RE, eds. The Systemic Sclerodermas and Related Disorders. 4th ed. Philadelphia, PA: WB Saunders; 2001:505–543
- ↵Crocq M. De l’ “ acrocyanose.” Semaine Med.1896;16 :298
- ↵Cherniack M, Clive J, Seidner A. Vibration exposure, smoking, and vascular dysfunction. Occup Environ Med.2000;57 :341– 347
- ↵Kenet G, Sadetzki S, Murad H, et al. Factor V Leiden and antiphospholipid antibodies are significant risk factors for ischemic stroke in children. Stroke.2000;31 :1283– 1288
- ↵Bhanji S, Mattingly D. Acrocyanosis in anorexia nervosa. Postgrad Med J.1991;67 :33– 35
- Copyright © 2003 by the American Academy of Pediatrics