Date sent: Fri, 11 Oct 2002 Source: Human Biology 74(4):615-20 Date: August 2002 URL: http://muse.jhu.edu/journals/hub/ http://muse.jhu.edu/journals/human_biology/v074/74.4brimacombe.html Birth Order and its Association with the Onset of Chronic Fatigue Syndrome -------------------------------------------------------------------------- Michael Brimacombe,1,2,5 Drew A. Helmer,3,5 and Benjamin H. Natelson1,4,5 1 Chronic Fatigue Syndrome Cooperative Research Center, University of Medicine and Dentistry, New Jersey Medical School, Newark, New Jersey 07103. 2 Department of Preventive Medicine, University of Medicine and Dentistry, New Jersey Medical School, Newark, New Jersey 07103. 3 Department of Medicine, University of Medicine and Dentistry, New Jersey Medical School, Newark, New Jersey 07103. 4 Department of Neurosciences, University of Medicine and Dentistry, New Jersey Medical School, Newark, New Jersey 07103. 5 War-Related Illness and Injury Study Center, Veterans Administration New Jersey Health Care System, East Orange, New Jersey 07018. Received 10 January 2002; revision received 9 May 2002. Abstract Chronic fatigue syndrome (CFS) is a medically unexplained illness that is diagnosed on the basis of a clinical case definition; so it probably is an illness with multiple causes producing the same clinical picture. One way of dealing with this heterogeneity is to stratify patients based on illness onset. We hypothesized that either the whole group of CFS patients or that group which developed CFS gradually would show a relation with birth order, while patients who developed CFS suddenly, probably due to a viral illness, would not show such a relation. We hypothesized the birth order effect in the gradual onset group because those patients have more psychological problems, and birth order effects have been shown for psychological characteristics. We compared birth order in our CFS patients to that in a comparison group derived from U.S. demographic data. We found a tendency that did not reach formal statistical significance for a birth order effect in the gradual onset group, but not in either the sudden onset or combined total group. However, the birth order effect we found was due to relatively increased rates of CFS in second-born children; prior birth order studies of personality characteristics have found such effects to be skewed toward first-born children. Thus, our data do support a birth order effect in a subset of patients with CFS. The results of this study should encourage a larger multicenter study to further explore and understand this relation. Keywords: chronic fatigue syndrome, illness onset, birth order --------------------------------------------------------------------------- Chronic fatigue syndrome (CFS) is a medically unexplained illness characterized by new onset of fatigue that produces some degree of disability as well as a wide variety of accompanying symptoms. Although many hypotheses have been posited regarding the cause of CFS (Natelson 2001), none has proven to be definitive. These hypotheses range from purely medical (i.e., a viral cause) to purely psychological (i.e., the problem is one of amplification of commonly occurring symptoms). Some years ago, a CFS patient Internet bulletin board was filled with electronic mail communications concerning the idea that there was a potential birth order effect in CFS. To our knowledge, no one has actually tested to see if such an order effect does exist. We can examine the existence of a birth order effect using our population of referred CFS patients, subject to the restrictions of the existing data and in relation to birth order in the general U.S. population. Birth order studies have been frequently used to look at psychological or psychosocial variables. Thus, the issue of whether CFS is a medical illness or a psychological state is critical in designing a birth order study. If CFS is simply an overreaction to symptoms commonly occurring in some people, then we might expect to find a birth order effect for the whole group of CFS patients. But because CFS is diagnosed based on a clinical case definition, it probably is not a single disorder but instead may be comprised of a number of disorders with different causes but a similar clinical picture. One technique recommended to reduce such heterogeneity is stratification into groups (Fukuda et al. 1994). One source of variability among patients with CFS has to do with illness onset. Many patients report that their illness develops to its full degree in 1-2 days with an acute influenza-like onset peaking in early winter (Zhang et al. 2000). In contrast, other patients report that it takes weeks to months for their illness to develop completely; this gradual onset group has significantly more comorbid psychiatric illness than the sudden onset group (DeLuca et al. 1997). We would not expect to find a birth order effect in individuals whose CFS follows an apparent viral infection. Alternatively, if psychological factors play a role in the gradual onset of CFS, one might see a birth order effect in this subgroup of patients. The purpose of this paper is to communicate evidence supporting this possibility. MATERIALS AND METHODS Our plan was to compare birth order data drawn from our own patient sample with that from a comparison population taken from the 1998 U.S. National Vital Statistics database. One of the reviewers of the original version of this manuscript had a concern that the birth order distribution using the 1998 database might differ from that in previous years. One suggestion to bypass this potential problem was to use an alternative reference population of birth order-that is, one based on the birth years of the sample CFS patients. Our sample varies in age from 27 to 52 with an interquartile range of 30 to 44. The birth years associated with that sample run from the early 1950s to the early 1970s. National Vital Statistics for that time period are far less accurate than current values. Birth order distributions for Caucasian women do exist for the years 1950, 1955, and 1959 onwards. This time frame, unfortunately, leaves out several sample birth years of interest. In addition, the existing recorded values for all the sample birth years are only estimates based on a 50% sample of births (the 1967 data was based on a 20%-50% sample). As such, they display random fluctuation over time due to sampling error and not necessarily due to actual changes in the underlying birth order distribution itself. Because of these shortcomings, we did not feel that use of these numbers would address the issue of potential temporal fluctuation in the actual birth order distribution for our sample. By using 1998 Vital Statistics data, we are employing a year for which there is full ascertainment. If we use, for example, the average birth order distribution for the years in which there is full ascertainment (1985-2000), there is no significant difference from 1998, and our results do not alter. Further, the greatest change in the birth order distribution for white women has been the decrease in higher values due to smaller families. We have avoided these higher values by restricting interest to birth orders less than 6 here. Thus, we feel use of the 1998 database is reasonable and appropriate for the research purposes of this report. As the ethnic distribution of our Center patient population is 94% Caucasian, we used Vital Statistics data drawn from Caucasians only (2000) as well as CFS Center data for Caucasians. Thus, the subject population with available birth order information was comprised of 219 patients who presented to the CFS Cooperative Center, located at the New Jersey Medical School's Doctors Office Center, and who fulfilled the published case definition for CFS (Fukuda et al. 1994). To further reduce heterogeneity (Natelson et al. 1995), these patients did not include any who had a psychiatric diagnosis occurring before the onset of their fatiguing illness. Comparisons of birth order for all CFS patients, sudden onset and gradual onset CFS groups, were examined, respectively, in regard to the U.S. reference population (see Table 1). We further examined the possibil- ity that gender may have played a role in some aspects of birth order or illness onset. Table 1. Number of People Falling within Each Birth Order (The Proportion of Total for Each Class Interval in Parentheses) ------------------------------------------------------------------------- Birth Order Gradual Onset Sudden Onset Population ------------------------------------------------------------------------- 1 30 (0.31) 32 (.34) 1252522 (0.41) 2 39 (0.40) 34 (0.36) 1032725 (0.33) 3 18 (0.19) 15 (0.16) 512186 (0.17) 4 5 (0.19) 7 (0.07) 188211 (0.06) 5 5 (0.05) 3 (0.03) 64535 (0.02) 6 - 2 (0.02) 25024 (0.01) 7 - 1 (0.01) 11166 (0.00) ------------------------------------------------------------------------- RESULTS There were no significant differences in age or education between the gradual onset and sudden onset groups (age: 38.2 p/m 9.5 [SD] vs. 36.1 p/m 8.5; education: 15.0 p/m 2.5 vs. 15.2 p/m 2.4). There was no significant difference in birth order distribution between the reference U.S. population of Caucasians only and the overall group of CFS patients (c^2=8.473; degrees of freedom [df]=6; p=0.21). There was also no difference from the group of patients with sudden onset (c^2= 0.306; df=4; p=0.98). Note that data were restricted to birth order less than 6 to avoid empty cells in the sample birth order distribution. However, a difference from the base population distribution was found for patients with gradual onset of CFS (c^2=8.047; df=4; p=0.09). The restriction of birth order less than 6 was maintained for comparability. Gender was not associated with birth order (exact test, p-value=0.83). Gender was also not associated with onset type (c^2=0.49, df=1, p=0.48), with females comprising 78.8% of the gradual and 75% of the sudden onset groups. Gender was also found to be unrelated to family size (c^2= 8.78, df=10, p=0.55). DISCUSSION The history of medicine has made clear that when an illness is diagnosed based on a clinical case definition, the delineated syndrome always has multiple causes. Thus the diagnosis of "dropsy" no longer exists in modern medicine, because it was replaced with specific diagnoses relating to the affected organ producing the clinical syndrome (Natelson 1990). This is the case too for CFS. In our prior work, we have stratified our patient sample based on the pattern of illness onset. Individuals whose illness developed over just a few days had a lower rate of having concurrent psychopathology than those whose illness developed gradually, over weeks to months (DeLuca et al. 1997). The sudden onset group reported illness onset that peaked in winter months, suggesting a viral cause (Zhang et al. 2000). We would not expect to find a birth order effect for this group. We did hypothesize that such an effect would be found, however, for the group with a gradual illness onset. The data reported here, although not fulfilling formal criteria for statistical significance, do lend support to our prediction. That we found the relation for the gradual onset group but not for the sudden onset group suggests two conclusions. First, as expected, CFS is a heterogeneous disorder with multiple causes producing the same end clinical picture, and, second, that the group with gradual illness onset has some factor related to birth order that plays a role in illness onset. The identity of that factor is not known, but it may have to do with personality characteristics of CFS patients-specifically, with their reporting themselves to be more action-prone (Van Houdenhove et al. 1995) and harder driving (Lewis et al. 1994) prior to the onset of their illness. Although we could find no study that evaluated birth order and these specific variables, available data do support a birth order effect for similar personality characteristics-namely, being dominant, self-confident, and more interested in being in control (Sulloway 1997). Because individuals with these personality characteristics are found to be firstborns rather than later-borns, we hypothesized that any birth order effect we might find would favor firstborns. The data did not support that specific hypothesis. Instead of finding an increased rate of firstborns in the gradual onset group, it was, if anything, reduced, and instead we found an increased rate of second-borns in that group. This intriguing finding will need to be confirmed and possible explanations explored with additional research. Note that gender plays a role in CFS. As we report here, CFS is an illness that predominantly strikes females. This may affect interpretation of our results, but given that approximately 50% of the U.S. population is female and birth order would be expected to be randomly distributed in regard to gender, the results reported here using the overall U.S. population do not lessen in relevance. One limitation of this study is the lack of information regarding family demographics and the socioeconomic status of parents of patients (Angst and Angst 1983). While family size did not differ significantly between the two CFS subgroups, it is possible that the families of patients with gradual onset illness differed in some important underlying demographic or psychosocial sense from CFS patients with sudden onset illness. It is possible that those differences rather than the presence of CFS are responsible for the difference detected in birth order. Also, the number of patients included is relatively small, reflecting the difficulty in recruiting patients with disabling fatigue. Nonetheless, the finding did support our a priori hypothesis, and thus it should be of interest to the medical and scientific community trying to understand this complex, often disabling illness. We hope this report will stimulate a multicenter study that can provide larger samples with sufficient power to detect a truly significant difference in birth order among patients with gradual onset chronic fatigue syndrome as suggested by this report. When such a study is done, it will be critical to pay attention to coincidental factors relating to family demographics or dynamics that might affect the outcome of such a study of birth order. ACKNOWLEDGMENTS This study was supported by the National Institutes of Health Center grant AI-32247. Literature Cited ---------------- Angst, C., and J. Angst. 1983. Birth Order: Its Influence on Personality. New York, NY: Springer Verlag. DeLuca, J., S.K. Johnson, S.P. Ellis et al. 1997. Sudden vs. gradual onset of chronic fatigue syndrome differentiates individuals on cognitive and psychiatric measures. J. Psychiatr. Res. 31:83-90. Fukuda, K., S.E. Straus, I. Hickie et al. 1994. The chronic fatigue syndrome: A comprehensive approach to its definition and study. Ann. Intern. Med. 121:953-959. Lewis, S., C.L. Cooper, and D. Bennett. 1994. Psychosocial factors and chronic fatigue syndrome. Psychol. Med. 24:661-671. Natelson, B.H. 1990. Tomorrow's Doctors: The Path to Successful Practice in the 1990s. New York, NY: Plenum Publishing Corporation. Natelson, B.H. 2001. Chronic fatigue syndrome. J.A.M.A. 285:2557-2559. Natelson, B.H., S.K. Johnson, J. DeLuca et al. 1995. Reducing heterogeneity in chronic fatigue syndrome: A comparison with depression and multiple sclerosis. Clin. Infect. Dis. 21:1204-1210. National Vital Statistics Report. 2000. Vol. 48. Washington, D.C.: National Center for Health Statistics. Sulloway, F.J. 1997. Birth order and personality. Born to Rebel: Birth Order, Family Dynamics, and Creative Lives, New York, N.Y.: Vintage Books, 55-79. Van Houdenhove, B., P. Onghena, E. Neerinckx et al. 1995. Does high 'action-proneness' make people more vulnerable to chronic fatigue syndrome: A controlled psychometric study. J. Psychosom. Res. 39:633-640. Zhang, Q, B.H. Natelson, J.E. Ottenweller et al. 2000. Chronic fatigue syndrome beginning suddenly occurs seasonally over the year. Chronobiol. Int. 17:95-99. -------- (c) 2002 by The Wayne State University Press