Secondary Logo

Journal Logo

Epidemiology and Social Science

Temporary Employment, Absence of Stable Partnership, and Risk of Hospitalization or Death During the Course of HIV Infection

Dray-Spira, Rosemary MD*; Gueguen, Alice PhD*; Persoz, Anne MSc; Deveau, Christiane MD; Lert, France PhD*; Delfraissy, Jean-François MD, PhD; Meyer, Laurence MD, PhDthe PRIMO Cohort Study Group

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: October 1st, 2005 - Volume 40 - Issue 2 - p 190-197
doi: 10.1097/01.qai.0000165908.12333.4e


Since the advent of highly active antiretroviral therapy (HAART), HIV infection has become a chronic disease in high-income countries and nonspecific HIV-related morbidity has been assuming a growing role in the overall morbidity and mortality of HIV-infected persons.1-7 In this new context, the global health status of HIV-infected persons has been shown to be largely influenced by determinants of non-HIV-related morbidity, including antiretroviral regimen, level of therapeutic adherence, comorbidity, health behavior-related exposure to other pathologic entities, and level of social support.8 Such determinants themselves largely depend on patients' socioeconomic conditions.8

Health benefits of recent advances in HIV infection management have been shown to differ between groups of patients independent of their clinical characteristics, with poorer health outcomes reported among ethnic minorities, female patients, youngest patients, injection drug users, and individuals without a stable partnership.2,3,9-12 Such social health inequalities are probably at least partly explained by socioeconomic differences in the distribution of classic determinants of HIV-infected patients' health status (eg, access to health care, treatment adherence). In the context of HIV disease as well as in the general population,13 however, a part of the association between social characteristics and health status may also exist that remains unexplained by classic health determinants, suggesting the existence of other mechanisms linking living conditions and health.

In France, access to care for HIV infection is universal. In the HAART era, HIV infection has been mostly managed through outpatient clinics, and inpatient admissions have been limited to the most serious cases. Thus, the indicator constituted by all-cause hospitalization or death is likely to represent a valid marker of HIV-infected patients' global health deterioration.

The aim of this study was to estimate the independent association between socioeconomic situation and the risk of all-cause hospitalization or death during the course of HIV disease among a population of HIV-infected patients followed from the time of primary infection in France.


Study Design

The French Agence Nationale de Recherche sur le SIDA (ANRS) EP8 PRIMO cohort is an ongoing, observational, prospective study conducted since November 1996 in 66 clinical settings delivering HIV care and located all over the French territory. Study design has been detailed elsewhere.14 In the participating hospitals, patients are eligible for enrollment if (1) they present during or soon after development of HIV-1 primary infection (symptomatic or not symptomatic), (2) they are antiretroviral naive, and (3) they give their written informed consent to participate. Primary HIV-1 infection is confirmed by an incomplete Western Blot test (ie, absence of anti-p68 and anti-p34 antibodies) or a positive p24 antigenemia with a negative or weakly negative enzyme-linked immunosorbent assay (ELISA) test or an interval of less than 6 months between the last negative and first positive ELISA test results. The study has been approved by the local Comite Consultatif de Protection des Personnes se pretant à la Recherche Biomedicale (Paris, France). After enrollment, participants are followed up through hospital outpatient visits at day 14, month 1 (M1), month 3 (M3), month 6 (M6), and semiannually thereafter.

Data Collection

At enrollment and at each scheduled visit, data are collected through a physician-administered standard questionnaire and patients are asked to answer a self-administered questionnaire focusing on sexual relations during the last 6-month period. At baseline, information is collected on sociodemographic characteristics (gender, country of birth, age, educational level, occupational position, employment status, and stable partnership in the last 6 months), characteristics of HIV disease and its management (transmission category; CD4 cell count; viral load; and antiretroviral regimen categorized as multitherapy, including a protease inhibitor [yes/no]), chronic comorbidity categorized as none, depression, or other (including hepatitis B or C virus, respiratory disease, diabetes, and high blood pressure), acute comorbidity categorized as none, mental disorder, or other (including sexually transmitted disease, liver-related disorder, and non-HIV-related infection), and risky health behaviors (tobacco consumption and ongoing injection drug use). At each follow-up visit, information on the occurrence of at least 1 inpatient admission during the period since the last cohort visit (“Has the patient been hospitalized since last visit? [yes/no]”) and on the characteristics of this hospitalization (date of entry, length of stay, and causes) is recorded. Moreover, updated data are collected on employment status and stable partnership, characteristics of HIV disease and its management (CD4 cell count, viral load, clinical stage, and antiretroviral regimen), level of treatment adherence (categorized as high, medium, low, or untreated), acute comorbidity, and tobacco consumption. Death notification is made by physicians.

To mitigate potential bias caused by sicker patients discontinuing stable employment or partnership because of their poor health status, as previously described,12 we assumed, according to a conservative principle, that once patients had reported stable employment or a stable partnership, they remained in this social situation until the end of follow-up.

Statistical Analysis

Analyses were restricted to the 319 patients who had attended at least the M6 visit at the cutoff point of December 31, 2002. We used all-cause hospitalization or death as the marker of patients' overall health status during the course of HIV infection. To ensure the quality of this marker, hospitalization related to causes that did not reflect deterioration in patients' health status (hospitalization for symptoms of primary HIV infection and for pregnancy, child birth, or therapeutic termination of pregnancy) were excluded, as were day hospitalizations, generally used in the French context of HIV care for advanced laboratory and/or radiologic investigations.

For each patient, data were split in as many “at-risk” observations as the number of 6-month periods between 2 outpatient visits, and the outcome variable (occurrence of at least 1 hospitalization or death [yes/no]) was created for each of these periods. Associations between updated characteristics of the social situation (employment status and stable partnership) reported at each observation's outset visit and the risk of hospitalization or death during the subsequent 6-month period of observation were assessed using univariate and multivariate logistic regression models based on generalized estimating equations (GEEs) of Liang and Zeger,15 assuming an exchangeable working correlation structure. We used covariates fixed over time (gender, country of birth, educational level, HIV transmission category, baseline ongoing injection drug use, tobacco consumption, baseline CD4 cell count and HIV viral load, and baseline chronic comorbidity) and dependent on time (age, CD4 cell count, HIV viral load, clinical stage, antiretroviral regimen, level of treatment adherence, and acute comorbidity). Moreover, because the risk of hospitalization or death is likely to vary over time since infection and the management of HIV primary infection has changed since 1996,16 updated time since enrollment and calendar period (dichotomized as ≤1999 and >1999) were systematically entered in the multivariate model as variables of adjustment.

Distribution of causes of hospitalization or death and lengths of hospital stay were described according to employment status and stable partnership at the time of the observation's outset visit. Statistical analyses were performed using Stata 7.0 (Stata Corporation, College Station, TX).


Sample Characteristics

As of December 31, 2002, the 319 enrolled patients had been followed during a median time of 2.5 years (range: 5.3 months to 6.5 years) and had attended a median of 5 semiannual scheduled outpatient visits after enrollment, representing a total of 1751 observations between 2 semiannual cohort visits. At the cutoff point time, 282 patients had attended the M12 visit, 198 the M24 visit, 140 the M36 visit, 86 the M48 visit, 57 the M60 visit, and 15 the M72 visit. Thirty six patients (11.3%) had attended their last cohort visit before the January 1, 2002, after a median follow-up time of 2.0 years (range: 5.8 months to 4.5 years), and were thus considered as lost to follow-up for this analysis. At enrollment, these 36 patients were comparable to the others except that they had a lower educational level (Table 1).

Major Sociodemographic, Clinical, and Biologic Characteristics at Enrollment of the 319 Patients of the PRIMO Cohort

Major patient sociodemographic, clinical, and biologic characteristics at enrollment are shown in Table 1. Patients were predominantly male (82.2%) and natives of France or another European country (89.4%). The median age at enrollment was 33.6 years (range: 15.1 to 72.9 years). Most of the patients had been infected with HIV through homosexual or bisexual (62.4%) or heterosexual (27.3%) contact. Most (n = 248 [77.7%]) were employed at enrollment, among whom 19 had a nonpermanent job, and a stable partnership was reported by 196 patients (61.3%). At enrollment, 13 patients (4.1%) had severe immunosuppression with a CD4 cell count less than 200/mm3. A prescription for HAART was initiated on the day of enrollment in 218 patients (68.3%). After 1 year of follow-up, 224 (79.4%) of the 282 patients who had attended the M12 visit were being prescribed antiretroviral therapy (95% HAART). Thirty-eight patients (11.9%) had at least 1 chronic disease in addition to HIV infection at enrollment. Smoking was reported by 167 patients (52.4%), and ongoing injection drug use was reported by 3 patients (0.9%).

Risk of Hospitalization or Death

Occurrence of at least 1 hospitalization between 2 consecutive cohort visits was reported for 109 of the 1751 observations. These 109 events occurred in 84 (26.3%) of the 319 cohort patients: for 63 subjects, the occurrence of at least 1 hospitalization between 2 cohort visits was reported once during follow-up; for 18 subjects, it was reported 2 times; for 1 subject, it was reported 3 times; and for 2 subjects, it was reported 4 times. Moreover, 3 deaths occurred during follow-up. Thus, a total of 112 observations with hospitalization or death reported were considered for analyses.

Figure 1 shows the yearly proportion of observations with hospitalization and/or death reported from enrollment to 6 years of follow-up. The 112 events occurred at a median time of 1.1 years after enrollment (range: 3.0 months to 6.0 years): 54 (53 hospital admissions and 1 death) occurred during the first year of follow-up, 25 (all hospital admissions) occurred during the second year, and 33 (31 hospital admissions and 2 deaths) occurred from the start of the third year onward.

Yearly proportion (and 95% confidence interval) of observations with hospitalization or death reported from enrolment to 6 years of follow-up among the 319 patients of the PRIMO cohort.

Associations Between Social Characteristics and Risk of Hospitalization or Death

Results of the univariate and multivariate analyses comparing the 112 observations with hospitalization or death reported with the 1639 observations without hospitalization or death are presented in Table 2. Univariate analysis suggested that these 112 observations had occurred more frequently among patients who were unemployed or had a nonpermanent job in comparison to those with stable employment as well as among patients without any stable partnership compared with those with a stable partnership. Employment status and stable partnership remained independently associated with the risk of hospitalization or death in multivariate analysis, with a significantly increased risk for patients with temporary employment compared with those with stable employment (adjusted odds ratio [OR] = 2.5, 95% confidence interval: 1.1 to 5.6) and for patients without a stable partnership compared with those with a stable partnership (OR = 1.6, 95% confidence interval: 1.0 to 2.7); patients who were unemployed tended to be at increased risk compared with those with stable employment (OR = 1.7, 95% confidence interval: 0.9 to 3.5). The other patient characteristics significantly associated with an increased risk of hospitalization or death in multivariate analysis included baseline CD4 count less than 200 cells/mm3, updated viral load greater than 100,000 copies/mL, chronic depression at baseline, acute mental disorders during follow-up, and ongoing injection drug use at baseline; the risk of hospitalization or death also tended to be higher in case of chronic comorbidity other than depression, and it was independently higher during the first year after enrollment compared with the period from the second year of follow-up onward (see Table 2).

Factors Associated With Risk of Hospitalization or Death Among the 319 Patients of the PRIMO Cohort*

Causes of Hospitalization or Death

A diagnosis was documented for 108 of the 109 hospital admissions and for all 3 deaths that occurred during follow-up. As shown in Table 3, major causes of hospitalization included complications of HIV infection or HAART (fever, asthenia, drug-related liver disease, renal lithiasis, gastrointestinal disease, or dermatosis) in 27 hospitalizations (25.0%) and mental disorders (mainly depressive symptoms) in 23 hospitalizations (21.3%). Other causes of hospitalization (mostly non-HIV-related infections and surgical reasons) were reported for 58 inpatient admissions (53.7%). Causes of the 3 deaths were tuberculosis, suicide, and lymphoma.

Cause of Hospitalization or Death and Length of Hospitalization According to Employment Status and Stable Partnership in the PRIMO Cohort

Length of Hospital Stay

Overall, the median length of stay during the 109 hospitalizations was 6 days (range: 1 day to 3 months). As shown in Table 3, the median length of hospital stay varied according to employment status, ranging from 5 days among those with stable employment to 6.5 days among those unemployed and 8 days among those with a temporary job.


To our knowledge, although social issues have been weighing more and more on HIV clinical practice in Western countries as the disease has become chronic and the epidemic has shifted toward more socially vulnerable populations, only a few studies have focused on the relations between employment status, stable partnership, and health status during HIV infection. Of particular interest in our study is that patients were enrolled at the early stage of primary infection and prospectively followed afterward, allowing for the documentation of events occurring during the whole course of the disease, including early after infection.

Our results show that in the context of our cohort of HIV-infected patients in the era of HAART, even after adjustment for acknowledged health determinants of HIV-infected patients, social conditions are independently associated with patients' health status during the first years of the disease. Patients with nonpermanent employment are at increased risk of hospitalization or death compared with those with stable employment, as are patients without any stable partnership compared with those with a stable partnership. This last finding is consistent with recent data reporting a higher rate of progression to AIDS or death among HAART-treated patients without a stable partnership in the Swiss HIV cohort study.12

Our finding of an association between social conditions and health status may be interpreted in 2 ways: these characteristics may constitute determinants but also consequences of changes in health status, with the sicker patients being less likely to stay in a stable partnership or employment. Thanks to the fact that employment and partnership status were measured before the outcome, to our conservative definition of these characteristics, and to systematic adjustment for time-dependent health status, the associations we show are unlikely to reflect an effect of health deterioration on employment and partnership status. As in any observational study, however, unmeasured confounding may have occurred, and causality can not definitely be established.17

Moreover, several considerations suggest that the strength of the associations between social characteristics and health status may be underestimated in the present study. The cohort design is probably biased toward selection of the most socially privileged patients in our study sample: to be enrolled in the PRIMO cohort, patients had to be diagnosed as infected with HIV at the early stage of primary infection and had to become involved in a long-term prospective follow-up, conditions that are likely to exclude the most socially fragile patients.18 Thus, the social heterogeneity observed in our study sample is likely to be lower than in the entire population of HIV-infected patients in France. Moreover, social differences in health care (eg, differences in outpatient visit frequency or in access to antiretroviral treatments) reported among wider populations of HIV-infected patients19 are less likely to happen in the context of such a cohort study.

Although dropout may have biased our estimations, the fact that patients lost to follow-up were comparable to those followed regarding baseline characteristics and experience of hospitalization during follow-up (30.6% vs. 25.8%) and that most of them had stable employment (83.3%) and were in a stable partnership (80.6%) at the last follow-up visit is reassuring.

Social conditions may affect the risk of hospitalization or death in several ways. In the present study, patterns of health care use may differ according to patients' social characteristics and social distress may constitute, independent of health status, an important motivation for inpatient admission among the most socially fragile patients. Causes and lengths of hospitalization did not show obvious differences according to patients' social characteristics (although the sample size allowed little power to detect such differences), suggesting that such “social” inpatient admissions do not explain the differences observed in the risk of hospitalization or death according to social characteristics.

Thus, the relations we found between social characteristics and health status may be explained by different processes. First, temporary employment and the absence of a stable partnership may themselves generate stress with direct physiologic effects on the neuroendocrine and immune systems, resulting in HIV disease progression. Such a mechanism has been demonstrated for other stressful life events (eg, bereavement).20-24 Second, as indirect consequences of such stress, temporary employment and the absence of a stable partnership may result in increased risky health behaviors (smoking, alcohol, or drug consumption), adverse psychologic effects, and poorer treatment adherence.25-29 In the PRIMO cohort, questions on alcohol use were included in the questionnaire only recently (in 1999) and thus could not be considered for this early analysis; in addition, drug use was rare. Thus, although multivariate analysis was adjusted for available data on health behaviors, a part of the observed relations between social characteristics and health status may be explained by differences in these unmeasured risky behaviors. However, mental disorders and poor treatment adherence, which were documented longitudinally, were not reported more frequently in case of temporary employment or the absence of a stable partnership. Third, the absence of a stable partnership may be a wider indicator of lack of social support and/or social isolation, which have been independently related to increased rates of mortality and morbidity in the general population.30-33 Fourth, nonpermanent employment may be associated with adverse working conditions (eg, high level of physical demand, low control over work pace and scheduling, increased tension and conflicts at work), which may have an effect on patients' health status.34 Temporary employment may also be the result of a health selection process previous to HIV infection in which patients with the poorest health status and/or the least healthy lifestyle have been selected.29

Unemployment has been shown to be associated with increased mortality in the general population.35,36 Among our cohort of HIV-infected patients, the increased risk of hospitalization or death associated with unemployment seems to be lower than that associated with nonpermanent employment. In our sample, frequencies of risky health behaviors, mental disorders, and poor treatment adherence are higher among unemployed patients than among those with stable employment, and these differences partly explain the increased risk of hospitalization or death observed among the unemployed patients in univariate analysis. Moreover, in France, unemployed HIV-infected patients may have access to disability benefits, which may result in less stress and higher incomes compared with patients with temporary employment, who have to face employment disruptions.37

Overall, more than a quarter of the patients have been hospitalized at least once during a median follow-up period of 2.5 years, a rate that seems to be high in this population of young, recently HIV-infected, largely HAART-treated, and mostly non-hepatitis C virus-coinfected patients. The persisting high risk of inpatient admission in the HAART era reported among ethnic minorities and injection drug users3,9,38 therefore also seems to occur among more socially privileged patients. The proportion of hospital admissions motivated by mental disorders was high (21.3%). The result indicating that the risk of hospitalization or death is higher during the first year of follow-up compared with the later period was unexpected. This finding was not attributable to a higher frequency of side effects related to the start of antiretroviral treatment or to psychologic distress in reaction to diagnosis during this period. As suggested by sociologic studies,39,40 the “biographical disruption” consecutive to a diagnosis of chronic illness may lead to a period of increased social and health vulnerability.

In conclusion, in the era of HAART, the risk of hospitalization is high from the first months of HIV disease. Early psychologic support should seek to prevent a portion of these hospitalizations occurring at the early stages of the disease. Adverse social conditions, namely, temporary employment and the absence of a stable partnership, constitute independent risk factors of poor health status as defined by all-cause hospitalization or death among HIV-infected patients. Further studies with more detailed data on patients' living conditions should provide a better understanding of the mechanisms of such social health inequalities in this specific context.


The authors thank all the patients and physicians who participated in the study and Ilham Iraqui and Nabil Saichi for monitoring.


1. Falusi OM, Pulvirenti J, Sarazine J, et al. HIV-infected inpatients in the HAART era: how do hepatitis C virus coinfected patients differ? AIDS Patient Care STDS. 2003;17:13-16.
2. Fleishman JA, Hellinger FH. Recent trends in HIV-related inpatient admissions 1996-2000: a 7-state study. J Acquir Immune Defic Syndr. 2003;34:102-110.
3. Gebo KA, Diener-West M, Moore RD. Hospitalization rates differ by hepatitis C status in an urban HIV cohort. J Acquir Immune Defic Syndr. 2003;34:165-173.
4. Klein MB, Lalonde RG, Suissa S. The impact of hepatitis C virus coinfection on HIV progression before and after highly active antiretroviral therapy. J Acquir Immune Defic Syndr. 2003;33:365-372.
5. Lewden C, Salmon D, Morlat P, et al. Causes of death among human immunodeficiency virus (HIV)-infected adults in the era of potent antiretroviral therapy: emerging role of hepatitis and cancers, persistent role of AIDS. Int J Epidemiol. 2005;34:121-130.
6. Martin-Carbonero L, Soriano V, Valencia E, et al. Increasing impact of chronic viral hepatitis on hospital admissions and mortality among HIV-infected patients. AIDS Res Hum Retroviruses. 2001;17:1467-1471.
7. Reisler RB, Han C, Burman WJ, et al. Grade 4 events are as important as AIDS events in the era of HAART. J Acquir Immune Defic Syndr. 2003;34:379-386.
8. Dray-Spira R, Lert F. Social health inequalities during the course of chronic HIV disease in the era of HAART. AIDS. 2003;17:283-290.
9. Floris-Moore M, Lo Y, Klein RS, et al. Gender and hospitalization patterns among HIV-infected drug users before and after the availability of highly active antiretroviral therapy. J Acquir Immune Defic Syndr. 2003;34:331-337.
10. Moore RD, Keruly JC, Chaisson RE. Differences in HIV disease progression by injecting drug use in HIV-infected persons in care. J Acquir Immune Defic Syndr. 2004;35:46-51.
11. The HIV Research Network. Hospital and outpatient health services utilization among HIV-infected patients in care in 1999. J Acquir Immune Defic Syndr. 2002;30:21-26.
12. Young J, De Geest S, Spirig R, et al. Stable partnership and progression to AIDS or death in HIV infected patients receiving highly active antiretroviral therapy: Swiss HIV cohort study. BMJ. 2004;328:15.
13. Goldberg M, Melchior M, LeClerc A, et al. Epidemiology and social determinants of health inequalities [in French]. Rev Epidemiol Sante Publique. 2003;51:381-401.
14. Desquilbet L, Deveau C, Goujard C, et al. Increase in at-risk sexual behaviour among HIV-1-infected patients followed in the French PRIMO cohort. AIDS. 2002;16:2329-2333.
15. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13-22.
16. Schiffer V, Deveau C, Meyer L, et al. Recent changes in the management of primary HIV-1 infection: results from the French PRIMO cohort. HIV Med. 2004;5:326-333.
17. Hernan MA. A definition of causal effect for epidemiological research. J Epidemiol Community Health. 2004;58:265-271.
18. Goldberg M, Luce D. Selection effects in epidemiological cohorts: nature, causes and consequences [in French]. Rev Epidemiol Sante Publique. 2001;49:477-492.
19. Wood E, Montaner JS, Chan K, et al. Socioeconomic status, access to triple therapy, and survival from HIV-disease since 1996. AIDS. 2002;16:2065-2072.
20. Leserman J. HIV disease progression: depression, stress, and possible mechanisms. Biol Psychiatry. 2003;54:295-306.
21. Leserman J, Petitto JM, Golden RN, et al. Impact of stressful life events, depression, social support, coping, and cortisol on progression to AIDS. Am J Psychiatry. 2000;157:1221-1228.
22. McEwen BS, Seeman T. Protective and damaging effects of mediators of stress. Elaborating and testing the concepts of allostasis and allostatic load. Ann NY Acad Sci. 1999;896:30-47.
23. Theorell T, Blomkvist V, Jonsson H, et al. Social support and the development of immune function in human immunodeficiency virus infection. Psychosom Med. 1995;57:32-36.
24. Uchino BN, Cacioppo JT, Kiecolt-Glaser JK. The relationship between social support and physiological processes: a review with emphasis on underlying mechanisms and implications for health. Psychol Bull. 1996;119:488-531.
25. Ferrie JE, Shipley MJ, Marmot MG, et al. Job insecurity in white-collar workers: toward an explanation of associations with health. J Occup Health Psychol. 2001;6:26-42.
26. Gordillo V, Del Amo J, Soriano V, et al. Sociodemographic and psychological variables influencing adherence to antiretroviral therapy. AIDS. 1999;13:1763-1769.
27. Ingram KM, Jones DA, Fass RJ, et al. Social support and unsupportive social interactions: their association with depression among people living with HIV. AIDS Care. 1999;11:313-329.
28. Kivimaki M, Vahtera J, Pentti J, et al. Factors underlying the effect of organisational downsizing on health of employees: longitudinal cohort study. BMJ. 2000;320:971-975.
29. Kivimaki M, Vahtera J, Virtanen M, et al. Temporary employment and risk of overall and cause-specific mortality. Am J Epidemiol. 2003;158:663-668.
30. Berkman LF, Leo-Summers L, Horwitz RI. Emotional support and survival after myocardial infarction. A prospective, population-based study of the elderly. Ann Intern Med. 1992;117:1003-1009.
31. Eng PM, Rimm EB, Fitzmaurice G, et al. Social ties and change in social ties in relation to subsequent total and cause-specific mortality and coronary heart disease incidence in men. Am J Epidemiol. 2002;155:700-709.
32. Hanson BS, Isacsson SO, Janzon L, et al. Social network and social support influence mortality in elderly men. The prospective population study of “Men born in 1914,” Malmo, Sweden. Am J Epidemiol. 1989;130:100-111.
33. Seeman TE, Kaplan GA, Knudsen L, et al. Social network ties and mortality among the elderly in the Alameda County Study. Am J Epidemiol. 1987;126:714-723.
34. Theorell T. Working conditions and health. In: Berkman LF, Kawachi I, eds. Social Epidemiology. New York: Oxford University Press; 2000:95-117.
35. Moser KA, Fox AJ, Jones DR. Unemployment and mortality in the OPCS Longitudinal Study. Lancet. 1984;2:1324-1329.
36. Sorlie PD, Rogot E. Mortality by employment status in the National Longitudinal Mortality Study. Am J Epidemiol. 1990;132:983-992.
37. Dray-Spira R, Lert F, Marimoutou C, et al. Socio-economic conditions, health status and employment among persons living with HIV/AIDS in France in 2001. AIDS Care. 2003;15:739-748.
38. Gardner LI, Klein RS, Szczech LA, et al. Rates and risk factors for condition-specific hospitalizations in HIV-infected and uninfected women. J Acquir Immune Defic Syndr. 2003;34:320-330.
39. Bury M. Chronic illness as a biographical description. Sociol Health Illn. 1982;4:167-182.
40. Bury M. The sociology of chronic illness: a review of research and prospects. Sociol Health Illn. 1991;13:451-468.


PRIMO Cohort Study Group

J. Beylot, P. Morlat, D. Malvy, M. Bonarek, and F. Bonnet (St. André, Bordeaux); C. Caulin, E. Badsi, J. Cervoni, and V. Vincent (Lariboisière, Paris); J. M. Molina and D. Ponscarme (St. Louis, Paris); A. P. Blanc and T. Allègre (Aix en Provence); J. F. Delfraissy, C. Goujard, and Y. Quertainmont (Bicêtre, Le Kremlin Bicêtre); F. Raffi, V. Reliquet, E. Billaud, and J. L. Esnault (Hôtel-Dieu, Nantes); F. Bricaire, C. Katlama, V. Zeller, H. Ait Mohand, C. Duvivier, and J. Ghosn (Pitié-Salpétrière, Paris); E. Rouveix, S. Morelon, and C. Dupont (A. Paré, Boulogne); J. Reynes, V. Baillat, and V. Lemoing (Montpellier); J. M. Livrozet, F. Jeanblanc, and P. Chiarello (E. Herriot, Lyon); R. Thomas, F. Souala, and C. Bouvier (Pontchaillou, Rennes); A. Cabié and S. Abel (Fort de France); J. L. Vildé, C. Jestin, and C. Jadand (Bichat, Paris); E. Pichard, P. Fialaire, and J. M. Chennebault (Angers); P. Henon, G. Beck-Wirth, and C. Beck (Emile Muller, Mulhouse); D. Sereni and C. Lascoux (St. Louis, Paris); S. Herson and A. Simon (Pitié-Salpétrière, Paris); B. Dupont and J. P. Viard (Necker, Paris); A. Devidas, P. Chevojon, and P. Kousignian (Corbeil); P. Massip and M. Obadia (Purpan, Toulouse); J. Beytout and C. Jacomet (Hôtel Dieu, Clermont-Ferrand); H. Aumaître, B. Delmas, and M. Saada (Joffre, Perpignan); P. Yeni, E. Bouvet, I. Fournier, P. Campa, and S. Abgrall (Bichat, Paris); A. Sobel, P. Lesprit, and A. S. Lascaux (H. Mondor, Créteil); H. Gallais, I. Ravaux, and C. Tomei (La Conception, Marseille); R. Verdon and M. Six (Caen); C. Trepo and C. Augustin-Normand (Hôtel-Dieu, Lyon); G. Pialoux, W. Rosembaum, L. Slama, and P. Mariot (Tenon, Paris); P. Morel and F. Timsit (St. Louis, Paris); B. Hoen and C. Drobacheff (St. Jacques, Besançon); M. Kazatchkine, P. Castiel, and D. Batisse (HEGP, Paris); D. Sicard, D. Salmon, and A. Brunet (Cochin, Paris); P. Galanaud, F. Boué, and J. Polo de Veto (A. Béclère, Clamart); P. Veyssier and D. Merrien (Compiegne); P. M. Girard and D. Samanon-Bollens (St. Antoine, Paris); M. Bentata and F. Rouges (Avicenne, Bobigny); J. P. Cassuto, C. Sohn, and E. Rosenthal (L'Archet, Nice); P. Dellamonica and S. Chaillou (L'Archet, Nice); J. M. Ragnaud and I. Raymond (Pellegrin, Bordeaux); P. Choutet, P. Nau, and F. Bastides (Tours); P. Canton and L. Boyer (Nancy); Y. Mouton and A. Dos Santos (Tourcoing); P. Chavanet and M. Buisson (Dijon); G. Dien, C. Daniel, and C. Devaurs (St.-Brieuc); Y. Redelsperger, B. Ponge, and L. Fournier (Melun); J. Laffay and A. Greder Belan (A. Mignot, Le Chesnay); I. Lamaury and A. Cheret (Pointe a Pitre); M. Gayraud and L. Bodard (IMM Jourdan, Paris); J. C. Imbert and O. Picard (St. Antoine, Paris); E. Oksenhendler and L. Gérard (St. Louis, Paris); G. Huchon and A. Compagnucci (Hôtel-Dieu, Paris); P. Lagarde and F. David (Lagny); Ph. Vinceneux and M. Bloch (L. Mourier, Colombes); B. Audhuy and N. Plaisance (Colmar); O. Bletry and D. Zucman (Foch, Suresnes); L. Bernard and J. Salomon (R. Poincaré, Garches); M. Chousterman and V. Garray (Intercommunal, Creteil); A. Regnier (Vichy); M. Uzan and F. Saint-Dizier (Ducuing, Toulouse); J. J. Girard (Loches); P. Moreau and O. Vaillant (Lorient); F. Grihon (Noyon); A. Lepretre (Eaubonne); D. Houlbert (Alençon); F. Caron and Y. Debab (Rouen); F. Trémolières and V. Perronne (Mantes la Jolie); A. Lepeu and B. Slama (Avignon); E. Brottier and L. Faba (La Rochelle); C. Miodovski (Paris); R. Armero and E. Counillon (Frejus); G. Guermonprez and A. Dulioust (Briis s/Forges); P. Boudon and D. Malbec (Aulnay S/Bois); O. Patey and C. Semaille (Villeneuve St. Georges); J. Deville, G. Remy, and I. Beguinot (Reims); and G. Gonzalez and F. Sanlaville (Sens).


HIV infection; hospitalization; death; socioeconomic factors; prospective study; primary infection

© 2005 Lippincott Williams & Wilkins, Inc.