The rates of spontaneous abortion declined: a retrospective cohort study from an industrialized area in Italy, 1996-2010

http://jer.sciedupress.com Journal of Epidemiological Research 2015, Vol. 1, No. 1 ORIGINAL ARTICLES The rates of spontaneous abortion declined: a retrospective cohort study from an industrialized area in Italy, 1996-2010 Fabio Parazzini1, 2 , Elena Ricci Giuseppe Bulfoni1 ∗1 , Sonia Cipriani1 , Francesca Chiaffarino1 , Matteo Malvezzi1, 3 , Mirella Di Martino1 , 1 Dipartimento Materno Infantile, Fondazione IRCCS Cà Granda, Ospedale Maggiore Policlinico, Milano, Italy Dipartimento di Science Cliniche e di Comunità, Università degli Studi di Milano, Milano, Italy 3 Istituto di Ricerche Farmacologiche Mario Negri; Dipartimento di Scienze Cliniche e di Comunità, Sezione di Statistica Medica e Biometria “GA Maccacaro”, Università degli Studi di Milano, Italy 2 Received: May 20, 2015 DOI: 10.5430/jer.v1n1p20 Accepted: July 8, 2015 Online Published: July 15, 2015 URL: http://dx.doi.org/10.5430/jer.v1n1p20 A BSTRACT Background: Spontaneous abortions (SA) were analyzed in an industrialized Italian region over a 15-year period to evaluate time trends. Retrospective cohort study. Methods: Since 1991, in Lombardy a standard form has been used to register all discharges from public or private hospitals. This study analyzed the Lombardy Region registry of hospital admissions in 1996–2010. Using the hospital discharge register, SA-related crude admission rates per 100 pregnancies in women aged 15 to 50 years, residing in Lombardy, were computed; age- and nationality-adjusted rates were also calculated. SA cases were identified searching the database for the ICD-9 codes for SA. SA rates/100 pregnancies (livebirths+SAs+induced abortions/2) in strata of age and calendar year were computed. Poisson regression analysis was used to test trends over time. Results: The crude SA rate was 10.72/100 pregnancies in 1996 and 10.45 in 2010. The corresponding age-standardized rate was 9.21. SA frequency was 4-5 times higher in women aged ≥ 40 years in comparison to women aged < 20 years. Age-adjusted SA rates decreased, over time, in non native Italian women. Conclusions: In this population, living in an industrialized area, SA frequency/100 pregnancies has slightly declined during the last 15 years. Key Words: Spontaneous abortion, Temporal trends, Frequency, Time trend, Gender medicine 1. I NTRODUCTION These pollutants affect the reproductive outcome of pregnancy immediately when they become prevalent contamiSpontaneous abortion (SA) is a common adverse outcome nants in a specific area or they are banned. An analysis of SA of pregnancy. Several epidemiological and laboratory stud- temporal trends in specific geographic areas can be useful to ies have linked the frequency of SA with exposure to envi- evaluate whether new factors that can affect the outcome of ronmental factors such as air pollution[1] and smoking.[2–4] ∗ Correspondence: Elena Ricci; Email: ed.ricci@libero.it; Address: Dipartimento di Science Cliniche e di Comunità, Università degli Studi di Milano, via Commenda 12-20122 Milan, Italy. 20 ISSN 2377-9306 E-ISSN 2377-9330 http://jer.sciedupress.com Journal of Epidemiological Research pregnancy negatively had a significant presence in a specific geographic area. Along the same line, the ban of specific pollutants (e.g. smoking) may affect the frequency of SA, as it seems to be affecting the rate of preterm births.[5] 2015, Vol. 1, No. 1 by Susser,[9] to avoid distortion due to high rates of induced abortion in population. Since the frequency of IA vary greatly between ad within populations, this correction is needed to restore the comparability of SA rates. Frequency of SA may also be affected by several factors such as maternal age[6] or the use (or lack thereof) of certain assisted reproduction techniques (ART) in the clinical practice, factors that are markedly changed in the last decades.[7] Information on SA trends in different populations is scant in the literature[1, 8] These numbers were obtained from the same database. SA Rates per 100 pregnancies in strata of age and nationality (native Italian/not native Italian) were computed. Rates were standardized for age (in quinquennia) and woman’s nationality (native Italian/non-native Italian) by the direct method, using the 2010 cases as standard. Poisson regression models, with calendar-year fitted as a continuous variable, were In this paper we have analyzed the temporal trends of SA in used to evaluate any linear trends in rates over the study peLombardy, a highly industrialized area, over the 1996-2010 riod between 1996 and 2010. SA rates and trend tests were period. calculated for age groups. 2. M ETHODS 3. R ESULTS A standard form is used to register all discharges from public or private hospitals in Lombardy. Among other data, information on hospital, patient age and nationality are reported. We have obtained the data for the 1996-2010 period; since 1996, data on hospital admissions that lasted less than 24 hours long were also recorded. The total number of pregnancies in the 1996-2010 period was 1,777,011 and SA-related hospital admissions were 188,233. SA cases were identified searching for codes 632, 634.X (X=1 to 9) of the International Classification of Diseases Ninth Revision (ICD-9). The denominator for estimating frequency of SA was the total number of pregnancies including births, SA and induced abortions (IA)/2, as suggested Table 1 shows the crude and standardized SA rates per 100 pregnancies in strata of calendar year and nationality. The crude SA rate was 11.91/100 pregnancies in 1996 and 11.34 in 2010. When considering the age standardized rates by nationality, in 1996 the age-adjusted rate was 14.04 in native Italian women and 11.43 in non native Italian ones. In both groups, these rates, tested with Poisson’s regression model, tended to decrease, being 11.75 and 10.19 respectively in 2010 (P <.0001). Table 1. SA rates/100 pregnancies (births+SA+induced abortions/2), crude, age-adjusted and age- and nationality-adjusted; 1996-2010, Lombardy, Italy Calendar year Certified number of SA Rate/100 pregnancies Crude Rate/100 pregnancies Age-adjusted: Native Italian Rate/100 pregnancies Age-adjusted: non Native Italian Rate/100 pregnancies Age- and nationality-adjusted 1996 11,627 11.91 14.04 11.43 13.36 1997 11,540 11.75 13.76 11.66 13.21 1998 11,877 11.94 13.81 11.66 13.25 1999 12,154 11.97 13.69 11.75 13.18 2000 12,101 11.51 13.25 10.54 12.54 2001 12,047 11.42 12.95 10.56 12.33 2002 12,416 11.68 13.11 10.61 12.46 2003 12,719 11.64 12.91 10.67 12.33 2004 13,164 11.81 12.92 11.17 12.46 2005 12,855 11.59 12.60 10.53 12.06 2006 13,224 11.62 12.51 10.42 11.97 2007 13,257 11.66 12.28 11.16 11.99 2008 13,279 11.47 12.01 10.66 11.66 2009 13,104 11.34 11.85 10.40 11.48 2010 12,869 11.34 11.75 10.19 11.34 Published by Sciedu Press 21 http://jer.sciedupress.com Journal of Epidemiological Research The age-nationality-adjusted estimates also declined between 1996 and 2010 (P <.0001). A significant time trend was detected by means of Poisson regression model, adjusting for age and nationality (P <.0001). The SA rates increased with age (see Figure 1), but declined in all age classes (P <.0001), with the exception of the youngest women (aged less than 20 years, P >.05). 2015, Vol. 1, No. 1 In Italy, SA rates increased between 1980 and 1995,[6] probably because of older maternal age and changes in the environment. From 1995 to 2006, the crude SA rates (SA/livebirth) did not show any significant increase.[13] However, the initial apparent rise was, at least in part, due to women becoming more aware of early spontaneous abortion, as also suggested by Lang and Nuevo-Chiquero.[8] In the interpretation of these findings, some potential confounders should be considered. Age is the main risk for SA. In this population, SA frequency was 4-5 times higher in women aged over 40 than in those aged < 20 throughout the whole studied period (19962010), which is consistent with the literature.[6, 14] In younger women we did not find any significant change, whereas in women aged more than 20 years (more likely to be having a planned pregnancy) a declining SA trend was observed. Overall age-adjusted rates tended to decrease, though in Lombardy, in the same period, the mean age at birth increased Figure 1. Age- and nationality-adjusted SA rates/100 from 30.6 in 1996 to 31.5 in 2008.[15] Similar results have pregnancies (births+SA+induced abortions/2), by age class been found in other European populations, as in Denmark,[16] among women residing in Lombardy, 1996-2010 where a prospective register linkage study found that fetal loss is high in women in their late 30s or older, irrespective of reproductive history. In a Swedish register study,[17] rising 4. D ISCUSSION The general results of this study showed that SA fre- mean age at pregnancy of women only partially accounted quency/100 pregnancies has slightly declined during the last for the SA increase seen in 1983-2003; the authors suggested several possible explanation, among which more sensitive 15 years. urine pregnancy tests and gynecologic examinations by ulBefore discussing the results, some potential limitations trasound. should be considered. An increased SA risk has been inconsistently reported after In Lombardy all hospital admissions in private and public the introduction of ART. In Lombardy the frequency of ART hospitals are registered by law in an administrative regional has been increasing over the last 15 years: the estimated data base, therefore our data should be considered totally number of ART cycles was 849/million inhabitants in 2005 representative of the cases of SA admitted in hospital in the and 998/million inhabitants in 2008; more than 25% of ART region. procedures are performed in women aged ≥ 40 years.[18] It is possible that some cases could be treated conservatively However we did not observed any increase in SA frequency (i.e. wait and see) in outpatient centers. This mode of treat- in women aged 30-39 or ≥40 years, thus supporting that no ment should be more common in the recent years, if at all, effect on SA risk exists, at least at a population level. since early diagnosis of SA increased. This potential bias should tend to underestimate the frequency of SA. With regard to the quality of diagnosis, for administrative reasons, all medical records are reviewed, and the diagnoses confirmed by local medical officers. It has been suggested that analysis of SA trends based on administrative records may underestimated the frequency of SA, however this effect seems to be constant over time.[10] Smoking is a well recognized risk factor for spontaneous abortion.[2, 3, 19] The prevalence of women smokers in fertile age has not changed substantially over the last decades in Italy and in Lombardy.[20] However, women planning a pregnancy or seeking advice during pregnancy are strongly advised to quit smoking: about 90% follow this advice, though most relapse after delivery.[21] Even secondhand tobacco smoke had been suggested as a risk factor for miscarriage: There is a methodological issue regarding the competing risk a recent cross-sectional analysis of data from the Women’s [22] of IA: SA rates in populations in which IA is frequent tend Health Initiative found that high level of lifetime secondto be lower. For this reason, we adjusted the number of IAs hand smoke exposure significantly increased the risk estimate in the denominator, as suggested by several authors.[9, 11, 12] for SA. It has been suggested that tobacco control legislation 22 ISSN 2377-9306 E-ISSN 2377-9330 http://jer.sciedupress.com Journal of Epidemiological Research may affect reproductive health.[5] This sample may offer the opportunity to analyze the effect of these laws, from a population prospective. In Italy smoke-free legislation was promulgated in 2003 and became effective in 2005. We did not observed any drop of SA rates after 2005, but it is possible that the slow decline was, at least partially, due to a new attitude towards smoking in public places already adopted before the smoking ban. Since the early 2000s many hospitals, public offices, restaurants and cinemas had banned smoking or had separate rooms for smokers and non smokers. In fact, a Health Ministry directive[23] regulated the matter, indicating that smoking was banned in most public places. So the effect of less exposure to passive smoke if any, should already be present before 2005. 2015, Vol. 1, No. 1 idents increased from 5.3% in 1996 to 24.2% in 2008.[13] This increasing proportion may also explain part of the lower rates observed in the later calendar period in the age-adjusted SA rates. However, the age-adjusted rates by country of origin declined over the 1996-2010 period in native Italian as well as in the non native Italian group. In conclusion, this analysis shows that in Lombardy, Italy, Southern Europe, notwithstanding the increasing age at delivery, SA rates were declining over the 1996-2010 period. This finding is important, as it shows that in an industrialized area, the risk of SA seemed not affected by potential introduction of new pollutants. C ONFLICTS OF I NTEREST D ISCLOSURE The Authors declare no conflict of interest. Another interesting finding of this analysis is the observation that non-native Italian women have a lower SA risk than naACKNOWLEDGEMENTS tive Italian women. This may be due to the fact that healthy F.P., E.R.: study conception and drafting of manuscript. S.C., young women are more likely to migrate that non healthy F.C., G.B.: data analysis and interpretation; M.M. And DM. ones, though generally health conditions are worse in non M.: interpretation and revision of manuscript. native women than in native ones. Moreover, the pregnancy awareness and therefore SA at early stage may be lower in This study was partially funded by Fondazione IRCCS Ca’ this group of women, leading to an underestimate of the SA Granda, Ospedale Maggiore Policlinico, in the framework of “Fondi Ricerca Corrente”. The authors thank Dr. Carlo Zocrate.[8] chetti, Lombardy Region, Health Directorate, for providing In Lombardy the proportion of deliveries from foreign resdata. R EFERENCES [1] Kumar S. Occupational, environmental and lifestyle factors associated with spontaneous abortion. Reprod Sci. 2011; 18: 915-930. http://dx.doi.org/10.1177/1933719111413298 [2] George L, Granath F, Johansson ALV, et al. Environmental Tobacco Smoke and Risk of Spontaneous Abortion. Epidemiol. 2006; 17: 500-505. PMid:16837826. http://dx.doi.org/10.1097/01.ed e.0000229984.53726.33 [3] Blanco-Munoz J, Torres-Sanchez L, Lopez-Carrillo L. Exposure to maternal and paternal tobacco consumption and risk of spontaneous abortion. Publ Health Rep. 2009; 124: 317-322. PMid:19320374. [4] Agnesi R, Valentini F, Fedeli U, et al. Maternal exposures and risk of spontaneous abortion before and after a community oriented health education campaign. Eur J Public Health. 2011; 21: 282-285. http://dx.doi.org/10.1093/eurpub/ckq073 [5] Cox B, Martens E, Nemery B, et al. Impact of a stepwise introduction of smoke-free legislation on the rate of preterm births: analysis of routinely collected birth data. BMJ. 2013; 346: f441. http://dx.doi.org/10.1136/bmj.f441 [6] Osborn JF, Cattaruzza MS, Spinelli A. Risk of spontaneous abortion in Italy, 1978-1995, and the effect of maternal age, gravidity, marital status, and education. Am J Epidemiol. 2000; 151: 98-105. http: //dx.doi.org/10.1093/oxfordjournals.aje.a010128 [7] Sullivan EA, Zegers-Hochschild F, Mansour R, et al. International Committee for Monitoring Assisted Reproductive Technologies (ICMART) world report: assisted reproductive technology 2004. Hum Published by Sciedu Press Reprod. 2013; 28: 1375-90. http://dx.doi.org/10.1093/hum rep/det036 [8] Lang K, Nuevo-Chiquero A. Trends in self-reported spontaneous abortions: 1970-2000. Demography. 2012; 49: 989-1009. http: //dx.doi.org/10.1007/s13524-012-0113-0 [9] Susser E. Spontaneous abortion and induced abortion: an adjustment for the presence of induced abortion when estimating the rate of spontaneous abortion from cross-sectional studies. Am J Epidemiol. 1983; 117: 305-8. PMid:6600879. [10] Buss L, Tolstrup J, Munk C, et al. Spontaneous abortion: a prospective cohort study of younger women from the general population in Denmark. Validation, occurrence and risk determinants. Acta Obstet Gynecol Scand. 2006; 85: 467-75. PMid:16612710. http: //dx.doi.org/10.1080/00016340500494887 [11] Figa Talamanca I, Repetto F. Correcting spontaneous abortion rates for the presence of induced abortion. Am J Public Health. 1988; 78: 40-2. http://dx.doi.org/10.2105/AJPH.78.1.40 [12] Hilden J, Modvig J, Damsgaard MT, et al. Estimation of the spontaneous abortion risk in the presence of induced abortions. Stat Med. 1991; 10: 285-97. http://dx.doi.org/10.1002/sim.4780100 211 [13] Parazzini F, Bulfoni G, Cipriani S, et al. I ricoveri ostetrici negli ospedali della Lombardia nel 2008. (Admissions for obstetric reasons in Lombardy hospitals, 2008). It J Gynecol Obstet. 2010; 22: 13-31. [14] Meeker JD, Missmer SA, Cramer DW, et al. Maternal exposure to second-hand tobacco smoke and pregnancy outcome among cou- 23 http://jer.sciedupress.com [15] [16] [17] [18] [19] 24 Journal of Epidemiological Research ples undergoing assisted reproduction. Hum Reprod. 2007; 22: 33745. PMid:17053002. http://dx.doi.org/10.1093/humrep/de l406 Parazzini F, Foschi R, Bertuccio P, et al. La salute della gravidanza: un quadro d’insieme. (Health in pregnancy: an overview). Riv Ost Ginecol Prat Med Perinat. 2010; 25: 12-39. Nybo Andersen AM, Wohlfahrt J, Christens P, et al. Maternal age and fetal loss: population based register linkage study. BMJ. 2000; 320: 1708-1712. http://dx.doi.org/10.1136/bmj.320.7251. 1708 Adolfsson A, Larsson PG. Cumulative incidence of previous spontaneous abortion in Sweden in 1983-2003: a register study. Acta Obstet Ginecol Scand. 2006; 85: 741-747. Italian Health Ministry. Ministero della Salute: Relazione del Ministro del Lavoro, della Salute e delle Politiche Sociali al Parlamento sullo stato di attuazione della legge contenente norme in materia di procreazione medicalmente assistita (legge 19 febbraio 2004, n.40, articolo 15). Attività anno 2008. (Report about implementation of law about assisted reproductive technology, 2008). http://www.iss.it/binary/rpma/cont/relazione_ PARLAMENTO_30giu2010.pdf US Department of Health and Human Services. Health Consequences of Tobacco Use Among Women. Women and Smoking: A Report of 2015, Vol. 1, No. 1 the Surgeon General. Rockville, MD: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office of Smoking and Health; 2001. p.177–450. [20] Gallus S, Tramacere I, Pacifici R, et al. Smoking in Italy 2008-2009: a rise in prevalence related to the economic crisis? Prev Med. 2011; 52: 182-183. http://dx.doi.org/10.1016/j.ypmed.2010.11 .016 [21] Charrier L, Serafini P, Giordano L, et al. Smoking habits in Italian pregnant women: any changes after the ban? J Public Health Policy. 2010; 31: 51-8. http://dx.doi.org/10.1057/jphp.2009.43 [22] Hyland A, Piazza KM, Hovey KM, et al. Association of lifetime active and passive smoking with spontaneous abortion, stillbirth and tubal ectopic pregnancy: a cross-sectional analysis of hystorical data from the Women’s Health Initiative. Tob Control. 2015; 24: 328335. http://dx.doi.org/10.1136/tobaccocontrol-2013-0 51458 [23] Italian Health Ministry. Interpretazione ed applicazione delle leggi vigenti in materia di divieto di fumo (Explanation and implementation of in force laws about smoking ban]. Directive no. 4, 28th March 2001. http://www.lavoro.gov.it/NR/rdonlyres/A90CB04 8-FB01-4DE5-84FC-EAA73F0A0B3E/0/CIRC4_2001SAN.pdf ISSN 2377-9306 E-ISSN 2377-9330