Cardiovascular risk profile in mothers of a portuguese birth cohort

PDSPPROGRAMA DOUTORAL EM SAUDE PÚBLICA UNIVERSIDADE DO PORTO FACULDADE DE MEDICINA Elisabete Cristina Macedo Alves Cardiovascular Risk Profile in Mothers of a Portuguese Birth Cohort Porto | 2012 Elisabete Cristina Macedo Alves Cardiovascular Risk Profile in Mothers of a Portuguese Birth Cohort Porto | 2012 Dissertação de candidatura ao grau de Doutor apresentada à Faculdade de Medicina da Universidade do Porto II III Art.º 48º, § 3º - “A Faculdade não responde pelas doutrinas expendidas na dissertação.” (Regulamento da Faculdade de Medicina da Universidade do Porto – Decreto-Lei nº 19337 de 29 de Janeiro de 1931) IV V Corpo Catedrático da Faculdade de Medicina do Porto Professores Catedráticos Efectivos Doutor Manuel Alberto Coimbra Sobrinho Simões Doutor Jorge Manuel Mergulhão Castro Tavares Doutora Maria Amélia Duarte Ferreira Doutor José Agostinho Marques Lopes Doutor Patrício Manuel Vieira Araújo Soares Silva Doutor Daniel Filipe Lima Moura Doutor Alberto Manuel Barros da Silva Doutor José Manuel Lopes Teixeira Amarante Doutor José Henrique Dias Pinto de Barros Doutora Maria Fátima Machado Henriques Carneiro Doutora Isabel Maria Amorim Pereira Ramos Doutora Deolinda Maria Valente Alves Lima Teixeira Doutora Maria Dulce Cordeiro Madeira Doutor Altamiro Manuel Rodrigues Costa Pereira Doutor Rui Manuel Almeida Mota Cardoso Doutor António Carlos Freitas Ribeiro Saraiva Doutor José Carlos Neves da Cunha Areias Doutor Manuel Jesus Falcão Pestana Vasconcelos Doutor João Francisco Montenegro Andrade Lima Bernardes Doutora Maria Leonor Martins Soares David Doutor Rui Manuel Lopes Nunes Doutor José Eduardo Torres Eckenroth Guimarães Doutor Francisco Fernando Rocha Gonçalves Doutor José Manuel Pereira Dias de Castro Lopes Doutor Manuel António Caldeira Pais Clemente Doutor António Albino Coelho Marques Abrantes Teixeira Doutor Joaquim Adelino Correira Ferreira Leite Moreira VI VII Professores Jubilados ou Aposentados Doutor Abel José Sampaio da Costa Tavares Doutor Abel Vitorino Trigo Cabral Doutor Alexandre Alberto Guerra Sousa Pinto Doutor Álvaro Jerónimo Leal Machado de Aguiar Doutor Amândio Gomes Sampaio Tavares Doutor António Augusto Lopes Vaz Doutor António Carvalho Almeida Coimbra Doutor António Fernandes da Fonseca Doutor António Fernandes Oliveira Barbosa Ribeiro Braga Doutor António Germano Pina Silva Leal Doutor António José Pacheco Palha Doutor António Luís Tomé da Rocha Ribeiro Doutor António Manuel Sampaio de Araújo Teixeira Doutor Belmiro dos Santos Patrício Doutor Cândido Alves Hipólito Reis Doutor Carlos Rodrigo Magalhães Ramalhão Doutor Cassiano Pena de Abreu e Lima Doutor Daniel Santos Pinto Serrão Doutor Eduardo Jorge Cunha Rodrigues Pereira Doutor Fernando de Carvalho Cerqueira Magro Ferreira Doutor Fernando Tavarela Veloso Doutor Francisco de Sousa Lé Doutor Henrique José Ferreira Gonçalves Lecour de Menezes Doutor José Augusto Fleming Torrinha Doutor José Carvalho de Oliveira Doutor José Fernando Barros Castro Correia Doutor José Luís Medina Vieira Doutor José Manuel Costa Mesquita Guimarães Doutor Levi Eugénio Ribeiro Guerra Doutor Luís Alberto Martins Gomes de Almeida Doutor Manuel Augusto Cardoso de Oliveira Doutor Manuel Machado Rodrigues Gomes Doutor Manuel Maria Paula Barbosa Doutora Maria da Conceição Fernandes Marques Magalhães Doutora Maria Isabel Amorim de Azevedo Doutor Mário José Cerqueira Gomes Braga Doutor Serafim Correia Pinto Guimarães Doutor Valdemar Miguel Botelho dos Santos Cardoso Doutor Walter Friedrich Alfred Osswald XIV XV Aos meus pais, às minhas irmãs e ao Hugo, porque me acompanharam ao longo de todo este percurso e nunca deixaram de acreditar em mim. XVI XVII Os meus agradecimentos' À Professora Ana Azevedo, pela confiança depositada em mim e pelo apoio nos momentos de insegurança e dúvida. Agradeço ainda a determinação e persistência com que me orientou e por me ter incutido o gosto pela Epidemiologia. Ao Professor Henrique Barros, por me ter dado a oportunidade de integrar a sua equipa e de aprender sempre mais. A todos os co-autores dos artigos que constituem esta tese, pelo seu importante contributo. A todas as pessoas envolvidas no projecto Geração XXI e, em especial, àquelas que integraram a equipa de reavaliação aos 4 anos, pelo profissionalismo e dedicação constantes. A todos os colegas do Departamento de Epidemiologia Clínica, Medicina Preventiva e Saúde Pública e do ISPUP, pelo companheirismo e bom ambiente de trabalho. À Bárbara, à Marta e à Helena pela partilha de experiências, aprendizagens, dúvidas e receios. À Rute pela nossa longa amizade e pelas palavras amigas e apoio moral nesta recta final. A todos os amigos que me acompanharam ao longo deste percurso, que me incentivaram a continuar, e que me levaram docinhos e miminhos. Aos meus pais, porque sempre me ensinaram a persistir e nunca desistir dos meus sonhos. À Cila e à Ana, as minhas manas, pela confiança nas minhas capacidades e pelo apoio incondicional. Ao Hugo por ser o meu ponto de abrigo e me fazer feliz todos os dias da nossa vida. XVIII 1 T ABLE OF CONTENTS Abstract 3 Resumo 9 Introduction 17 Cardiovascular disease 19 Cardiovascular risk 20 Cardiovascular disease in women 23 Women’s reproductive health 28 Pregnancy and cardiovascular risk 29 Pregnancy: a stress test for maternal cardiovascular function 29 Pregnancy: opportunity for health promotion and disease prevention 34 Aims 39 Research methods 43 Paper I. Medical record review to recover missing data in a Portuguese birth cohort: agreement with self-reported data collected by questionnaire and inter-rater variability 53 Results 65 Paper II. Prevalence of self-reported cardiovascular risk factors in Portuguese women: a survey after delivery 67 Paper III. Impact of risk factors on hypertensive disorders in pregnancy, in primiparous and multiparous women 81 Paper IV. Fetal sex modifies the effect of gestational hypertensive disorders on blood pressure of mothers and children at 4 years 103 Paper V. Pregnancy and afterwards: opportunities to modify the course of the smoking epidemic in a female population at high risk. 131 Paper VI. Cardiovascular risk profile of mothers of a Portuguese birth cohort: a survey 4 years after delivery 161 General discussion 189 References 195 2 3 ABSTRACT 4 5 Cardiovascular diseases (CVD) result from a complex interplay of genetic, developmental, environmental and behavioural factors. CVD mortality rates increase with the number of risk factors to which an individual is exposed, supporting that the risk for CVD depends on the interaction among those factors to complete a sufficient causal mechanism. In young women, pregnancy puts a physiological stress on the body that can unmask an underlying propensity for chronic disease. Additionally, pregnancy is often regarded as a good opportunity for health promotion and disease prevention, due to the strong motivation of mothers to protect the health of the unborn baby. This thesis aims to study the cardiovascular risk profile of mothers of a Portuguese birth cohort (Geração XXI), before and during pregnancy, and 4 years after delivery. To answer this main question, the following 5 specific objectives were defined and pursued, resulting in 5 papers: 1. To estimate the prepregnancy prevalence of five major cardiovascular risk factors (overweight/obesity, smoking, hypertension, dyslipidemia and diabetes mellitus), and to describe their distribution by age, gravidity and indicators of socioeconomic position (SEP). 2. To assess the impact of age, education, family history of cardiovascular disease, prepregnancy BMI and pregnancy weight gain on hypertensive disorders in pregnancy, in primiparous and multiparous mothers. 3. To assess the extent to which gestational hypertensive disorders lead to higher blood pressure in women and their offspring as early as 4 years after birth, and to assess the effect of fetal sex on these consequences. 4. To assess the frequency and determinants of maternal smoking cessation during pregnancy and of the sustainability of smoking cessation 4 years after delivery. 5. To estimate the prevalence of established cardiovascular risk factors (smoking, low fruit and vegetables intake, sedentariness, general overweight/obesity, abdominal obesity, hypertension, dyslipidemia and diabetes mellitus), 4 years after delivery, and to describe their distribution by age, gravidity and indicators of socioeconomic position (SEP). The birth cohort Geração XXI was assembled between 2005 and 2006 at all 5 public maternity units covering the metropolitan area of Porto, Portugal. Trained interviewers invited the mothers to participate, after delivery. Of the invited mothers, 91.4% accepted to participate. A total of 8647 infants, corresponding to 8495 mothers, 12 A coorte de nascimento Geração XXI foi recrutada entre 2005 e 2006, nos 5 hospitais públicos com maternidade da área metropolitana do Porto, Portugal. Após o parto, inquiridores treinados convidaram as mães a participar. De entre as mães convidadas, 91,4% aceitaram participar. Um total de 8647 crianças e as respetivas 8495 mães foram incluídas na coorte. Através de uma entrevista presencial às mães, foram recolhidas informações sobre as características demográficas e socioeconómicas, história pessoal e familiar de doença, história ginecológica e obstétrica, cuidados pré-natais e estilos de vida. As mães e as crianças realizaram uma avaliação antropométrica. Os mesmos inquiridores treinados reviram também os processos clínicos na altura do parto, para recolher informações relativas aos cuidados pré-natais, à ocorrência de complicações durante a gravidez, e às características do parto e neonatais. Para recuperar informações em falta nos questionários, revimos os processos clínicos obstétricos das mães (ARTIGO I), por 2 revisores treinados. A concordância entre os dados do questionário e dos registos obstétricos para a história pessoal de diabetes foi altamente concordante, enquanto para a de hipertensão foi moderadamente concordante. Em geral, as discrepâncias encontradas para o peso e a altura indicavam valores mais elevados nos registos clínicos do que nos questionários, o que resultou em divergências na classe de índice de massa corporal em 10,3% das mulheres. A informação recolhida foi extremamente consistente entre os dois revisores, com máxima concordância para a diabetes gestacional e o peso ao nascimento. Aos 4 anos de idade das crianças, entre 2009 e 2011, todas as mães e os respetivos filhos foram convidados a participar na reavaliação da coorte, composta por uma entrevista e avaliação física. Caso recusassem, eram convidadas a responder a uma entrevista telefónica. Do total, 86,2% das crianças e 84,2% das mães foram reavaliadas. Durante a entrevista presencial, a informação foi recolhida por inquiridores treinados, utilizando dois questionários estruturados. O questionário da mãe continha informação sobre as suas características demográficas, história pessoal e familiar de doença, história ginecológica e obstétrica, e estilos de vida. O questionário da criança incluía informações relativas às características socioeconómicas parentais, e sobre o desenvolvimento, saúde e hábitos da criança. Foi realizado um exame físico às mães e crianças, incluindo avaliação antropométrica e medição da pressão arterial, e colheita de uma amostra de sangue em jejum. Na entrevista por telefone também se recolheu informações relativas à saúde da mãe e da criança, ainda que com um número mais restrito de questões. 13 Nesta amostra de mulheres Portuguesas, 21,3% tinham excesso de peso e 8,8% eram obesas, 26,6% fumavam e 11,2% eram ex-fumadoras, antes de engravidar (ARTIGO II). A prevalência de hipertensão arterial, dislipidemia e diabetes mellitus era de 1,7%, 1,7% e 0,6%, respectivamente, com uma clara tendência de agregação. A prevalência de todos os fatores de risco cardiovascular, com a exceção do tabagismo, aumentou com a idade e com o índice de massa corporal. A escolaridade e o rendimento associaram-se inversamente com a prevalência de excesso de peso. As fumadoras eram mais novas, com um índice de massa corporal mais adequado e tinham uma posição socioeconómica mais baixa. Globalmente, as complicações hipertensivas afetaram 4,6% das gravidezes de feto único, e estavam associadas a idade mais avançada, menor escolaridade, história familiar de doença cardiovascular e excesso de peso antes e durante a gravidez, de forma similar em mulheres primíparas e multíparas (ARTIGO III). Aproximadamente 50% dos casos de complicações hipertensivas na gravidez em mulheres primíparas e 70% em multíparas foram atribuíveis ao efeito conjunto de idade superior a 34 anos, escolaridade inferior a 12 anos, história familiar de doença cardiovascular, excesso de peso/obesidade e ganho de peso excessivo durante a gravidez. Quatro anos após o parto, as complicações hipertensivas gestacionais associaram-se a aumentos significativos na pressão arterial sistólica e diastólica da mãe (ARTIGO IV). O risco de hipertensão arterial nas mães que desenvolveram uma complicação hipertensiva durante a gravidez, comparativamente com as que tiveram uma gravidez sem hipertensão, foi cerca de 6 vezes superior nas que deram à luz uma criança do sexo feminino e 3 vezes maior nas que deram à luz uma criança do sexo masculino. Adicionalmente, a pressão arterial sistólica e diastólica aos 4 anos de idade foi significativamente superior nos rapazes cujas mães tiveram uma complicação hipertensiva durante a gravidez, enquanto não foi detetado nenhum efeito nas raparigas. Apesar de quase metade das fumadoras terem parado de fumar durante a gravidez, aproximadamente dois terços destas retomaram o consumo 4 anos após o parto (ARTIGO V). A cessação tabágica durante a gravidez foi mais frequente em mulheres primigestas, que viviam com o marido/companheiro, com escolaridade mais elevada, com excesso de peso ou obesas, com cuidados pré-natais adequados, que começaram a fumar mais tardiamente e que fumavam uma menor quantidade de cigarros diária. Aquelas que mantiveram a cessação tabágica eram mais velhas, e mais frequentemente primigestas, viviam com o marido/companheiro 4 anos após o parto, tinham um menor consumo de tabaco antes de engravidar, amamentaram 52 14 semanas ou mais, voltaram a engravidar após a gravidez índice e os seus filhos tinham mais frequentemente um diagnóstico médico de asma ou rinite. Na reavaliação da coorte, 25,3% das participantes eram fumadoras, 50,7% consumiam menos de três porções de frutas e vegetais por dia, e 81,3% não praticavam qualquer tipo de exercício físico de lazer (ARTIGO VI). Adicionalmente, 31,4% apresentavam excesso de peso, 21,3% eram obesas e 31,8% possuíam obesidade abdominal. Relativamente às comorbilidades cardiometabólicas, 8,7% das mulheres eram hipertensas e, a prevalência de dislipidemia e diabetes mellitus era de 18,5% e 0,9%, respectivamente. A presença de pelo menos um fator de risco de cada um dos três grupos (estilos de vida adversos, adiposidade e comorbilidades cardiometabólicas) foi observada em 17,0% das mulheres. Todos os fatores de risco estavam associados com o desemprego e com níveis de escolaridade e rendimento mais baixos. Em conclusão, as mulheres jovens Portuguesas apresentam um risco elevado de doença cardiovascular, antes e após a gravidez. A agregação de fatores de risco cardiovascular, assim como as elevadas prevalências de estilos de vida adversos e de adiposidade, mostram um perfil de risco cardiovascular desfavorável em jovens mães, desde o período pré-concepcional. A gravidez induz profundas alterações na saúde da mãe, desmascarando riscos prévios e/ou uma propensão acrescida para o desenvolvimento de doenças cardiovasculares. As características prévias à gravidez explicam uma elevada proporção de complicações hipertensivas durante a gravidez, com o excesso de peso antes e durante a gravidez a terem uma grande contribuição, particularmente nas mulheres primíparas. Adicionalmente, 4 anos após uma gravidez de feto único, agravada por uma complicação hipertensiva da gravidez, em mulheres primíparas, as mães já possuem valores elevados de pressão arterial e mais frequentemente apresentam hipertensão arterial crónica, sendo este efeito mais acentuado quando o feto é do sexo feminino. Relativamente às crianças, foi observado um efeito diferencial de acordo com o sexo da criança, o que suporta a hipótese da existência de causas heterogéneas para a hipertensão na gravidez e que o desenvolvimento uterino coloca os rapazes em maior risco no futuro. Durante a gravidez cerca de metade das mulheres parou de fumar. Contudo, 4 anos após o parto dois terços retomaram os hábitos tabágicos, revelando que esta mudança comportamental tem pouco impacto a longo prazo na trajetória tabágica das mulheres. 15 O elevado risco de doença cardiovascular descrito em mulheres jovens e aparentemente saudáveis, que tiveram pelo menos uma gravidez que resultou num nado-vivo, reforça a necessidade de implementar estratégias coerentes e eficazes de promoção de saúde e de prevenção de doença nas fases precoces da vida adulta, de forma a otimizar a saúde atual e futura das mulheres. 16 17 INTRODUCTION 18 19 C ARDIOVASCULAR DISEASE Cardiovascular diseases (CVD), defined as diseases of the heart and circulatory system, are the leading cause of death and morbidity worldwide (1), accounting for nearly 30% of deaths (2). In Europe, CVD are responsible for almost half of deaths (54% in women and 43% in men) (3). Since 1970, a continuous decrease in total cardiovascular mortality rates has been observed in men and women in Western Europe. In fact, between 1970 and 2000, total cardiovascular mortality decreased on average 50% and 60%, corresponding to an average annual decline of 1.8% and 2.0%, in men and women, respectively (4). Two main factors have contributed to this decline: decrease in incidence due to substantial reductions in the prevalence of some major cardiovascular risk factors and decrease in case-fatality due to improvements in treatments, which become more effective and widely used (5). However, despite the decline of ageadjusted cardiovascular death rates during the last decades (4, 6), coronary heart diseases and stroke remain the two most common causes of death in Europe (3). In Portugal, diseases of the circulatory system accounted for 32.2% of all deaths in 2006; 37.3% in women and 27.6% in men (7). The total CVD mortality rates in Portugal are higher than the European average, with cerebrovascular diseases being the single most important cause of death in Portugal (8). Beyond their contribution to mortality, CVD also has a large contribution to morbidity outcomes. Morbidity from CVD is more difficult to quantify than mortality, since at present there is no routinely updated source of CVD morbidity data covering the European population (3). Additionally, there are many different measures of morbidity which difficult the comparison of data. The disability adjusted life years (DALYs) lost, which is an aggregate measure of years of life lost due to premature death and years of healthy life lost to disability, is used as an indicator of the burden of the disease. The estimates provided by the World Health Organization (WHO) Global Burden of Disease project indicates that over 34 million DALYs (23% of all DALYs) are lost each year due to CVD, in Europe (9). Regarding more developed European countries, it is estimated that 17% of all DALYs lost are due to CVD, which represent the largest cause of disability after neuropsychiatric disorders. In less developed European countries, CVD were the leading cause of DALYs lost. 20 Moreover, CVD has major economic costs for Europe (3, 10). In 2006, CVD cost the health care systems of the European Union just under €110 billion, which represents around 10% of the total health care expenditure. However, looking only at the direct costs to the health care systems grossly underestimates the true cost of CVD, since production losses from death and illness in those of working age and from the informal care of people with the disease contribute greatly to the overall financial burden. Production losses due to mortality and morbidity associated with CVD cost almost €41 billion in the European Union, with around two-thirds of this cost due to death and one-third due to illness in those of working age. Overall, CVD is estimated to cost the European Union economy €192 billion a year, of which 57% is due to direct health care costs, 21% to productivity losses and 22% to the informal care of people with CVD (3). Similarly, in Portugal, 10% of the total health care expenditure is spent on CVD, with 69%, 15% and 16% of the overall cost of CVD due to direct health care costs, productivity losses and the informal care of people with CVD, respectively (3). Due to the extremely high human and economic impact of CVD (3, 6, 10), it is important to descriptively quantify the distribution of cardiovascular risk factors in order to monitor trends over time and guide preventive strategies to reduce the future burden of disease, by identifying groups at higher risk and understanding factors that contribute to the exposure. C ARDIOVASCULAR RISK CVD reflect a combination of genetic, developmental, environmental and behavioural factors (6, 11). The modifiable nature as well as the considerable impact of lifestyles on the development of CVD (12) emphasize the importance of investing in primary prevention. CVD result of a complex causal chain, with different roots in a multifaceted sequence of events over time, consisting of socioeconomic factors, environmental and community conditions, and individual behaviour. Figure 1 outlines the causal chain of ischaemic heart disease (2). Some elements in the chain, such as high blood pressure or cholesterol, act as a relatively direct cause of the disease, while others risks, located further back in the causal chain, act indirectly through intermediary factors. These risks include physical inactivity, alcohol, smoking or fat intake. Less 21 certainty can be attributed to the causal role of other risk factors, such as education and income. However, modifying these background causes is more likely to have amplifying effects, since such modifications have the potential to yield fundamental and sustained improvements to health. Figure 1. The causal chain. Major causes of ischaemic heart disease are shown. Arrows indicate some (but no all) of the pathways by which these causes interact (2). Worldwide, hypertension, dyslipidemia, diabetes, obesity, physical inactivity, poor diet with insufficient fruit and vegetables intake, excessive alcohol consumption, smoking and stress are responsible for approximately 90% of all cases of acute myocardial infarction in both sexes (12). According to the up-to-date WHO Report on health risks (2) these risks factors together, except stress, account for 61% of healthy years of life lost due to CVD and 61% of cardiovascular deaths. Among European populations, CVD is mainly attributable to cigarette smoking, high blood pressure, high serum cholesterol, diabetes, overweight/obesity and unhealthy diet (13). They are designated major risk factors for three reasons: their high prevalence in populations, their strong impact on coronary risk; and their preventability and reversibility, primarily by safe improvements in population lifestyles (13). The relation between these risk factors and CVD seem to be independent, strong, continuous and graded, and the modification of these risk factors can result in substantial reduction in mortality (5). In the Portuguese population, central obesity, defined through the waist-to-hip ratio, was the risk factor with the highest impact in acute myocardial infarction, being 28 W OMEN ’ S REPRODUCTIVE HEALTH Reproductive health is a state of complete physical, mental and social wellbeing, and not merely the absence of disease or infirmity, addressing the reproductive processes, functions and system at all stages of life (8, 58). Therefore, it implies that people are able to have a responsible, satisfying and safe sex life and that they have the capability to reproduce and the freedom to decide if, when and how often to do so. Implicit in this concept are the right of men and women to be informed and to have access to safe, effective, affordable and acceptable methods of fertility regulation of their choice, and the right of access appropriate health care services that will enable women to go safely through pregnancy and childbirth and provide couples with the best chance of having a healthy infant (8, 58). In this context, preconceptional, prenatal and postnatal care are critical components in the continuum of care of women and newborns (59). Women’s health in Portugal has experienced a huge overall improvement since the late 70s and the implementation of the National Health System, which ensures all citizens nearly free access to primary care centers and public hospitals (60). The dramatic improvements in perinatal health are worth noticing. In 2004, the maternal mortality rate was 7.7 per 100,000 live births, very similar to the Europe mean of 6.6 per 100,000 live births (61). Also, between 1996 and 2008, the infant mortality declined from 6.9 to 3.3 per 1000 live births, the neonatal mortality from 4.2 to 2.1 per 1000 live births and the perinatal mortality from 8.4 to 4.0 per 1000 live births (7). These declines during the last decades are major successes of Portuguese perinatal health (62). Pregnancy and childbirth are physiological events, but they carry risks that can be reduced by health-care interventions such as the provision of family planning and maternity care and access to safe abortion care (63). In women, preconceptional and prenatal care can provide the opportunities for regular risk assessment and prevention of pregnancy-related co-morbidities. Postpartum care is important for detecting and treating infections and other conditions, including postpartum depression, and for providing advice on family planning. This continued care throughout pregnancy, childbirth and the postpartum period is essential for health promotion at young ages. 29 P REGNANCY AND CARDIOVASCULAR RISK Pregnancy is a natural life event, in a woman’s life, but implies some psychological and physiological adaptations. It constitutes a stress test for maternal cardiovascular function (64), due to the occurrence of several adaptive changes in the woman’s body that facilitate foetal growth. At the same time, it is a good opportunity for health promotion and disease prevention (65), since mothers tend to be strongly motivated to adopt healthier lifestyles in order to protect the health of the unborn baby (65, 66). Pregnancy: a stress test for maternal cardiovascular function In healthy pregnancies, hemodynamic changes, hyperlipidaemia, insulin resistance, up-regulation of inflammatory markers and plasma volume expansion take place in women’s physiology to meet demands of the rapidly developing foetus (67, 68). During pregnancy, there is a fall of approximately 10 mmHg in blood pressure until 22 to 24 weeks (69). This reflects the marked reduction in total peripheral vascular resistance, which decreases by 25%, while cardiac output is increased by as much as 50% (68, 70, 71). The fall in blood pressure may obscure the diagnosis of pre-existing mild hypertension, particularly in young women who do not have blood pressure measured regularly before pregnancy. In the third trimester, the blood pressure gradually increases and may normalize to pre-pregnant values by term (69). Immediately after delivery, blood pressure usually falls, but increases again over the first five postnatal days (71). Even women whose blood pressure was normal throughout pregnancy may experience transient hypertension in the early postpartum period, probably reflecting a degree of vasomotor instability (69). Plasma and red cell volumes also expand in normal pregnancy by approximately 40% and 25%, respectively. These changes begin as early as the fourth week of gestation and peak around the 28th week (72, 73). The progressive rise in plasma and blood volume are likely adaptations, via renal sodium retention, to the vasodilatation and fall in blood pressure (74, 75). 30 Normal pregnancy is a state of insulin resistance, with a doubling in fasting insulin concentrations. This is due to changes in maternal hormonal and metabolic factors related to the placenta, adipose tissue, and the growth hormone axis (76, 77). The increased insulin resistance reaches a maximum in the third trimester, and improves following delivery (78, 79). During pregnancy, there are also changes in lipid profile, with a particular increase of around 300% in triglyceride levels and a 25–50% increase in total cholesterol (80). This gestational hyperlipidaemia fulfils the physiological role of supplying both cholesterol and triglyceride to the rapidly developing foetus (81, 82). Additionally, pregnancy is associated with a generalised maternal inflammatory response (83, 84) and with an overall state of hypercoaguability, due to increases in the levels of several coagulation factors (85). It is likely that this state of hypercoagulability may serve to limit life-threatening bleeding at delivery but there is an increased risk of thromboembolism associated with pregnancy (86). Altogether, these changes in maternal physiology allow the cardiovascular system to adjust to the physiological demands of the foetus while maintaining maternal cardiovascular integrity (67). However, in some women, this normal adaptive response is greatly exaggerated and can lead to prenatal complications, such as recurrent miscarriage, congenital malformations, gestational hypertensive disorders, gestational diabetes mellitus and venous thromboembolism. Additionally, data linking the maternal vascular, metabolic, and inflammatory complications of pregnancy with an increased risk of vascular disease in later life, have been increasing (64). Moreover, the marked pregnancy-induced hemodynamic alterations put a physiological stress on the body that have a profound effect on future health. Hypertensive disorders in pregnancy have been shown to be predictors for hypertension and CVD events (87, 88), with women with placental complications (89), poor foetal growth or intrauterine death considered to be at the greatest risk (90). Further, an impaired glucose tolerance during pregnancy and gestational diabetes constitute female specific risk factors for the development of diabetes and the metabolic syndrome in relatively young women (91, 92). Hypertensive disorders in pregnancy are major complications of pregnancy, increasing the risk of adverse obstetric and perinatal outcomes (93-95). They include chronic and gestational hypertension, preeclampsia and eclampsia, with the latter three entities representing generally transient hypertensive conditions with onset during pregnancy (96). According to the National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy (96), chronic hypertension is defined as systolic blood pressure ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg, 31 diagnosed before pregnancy or before the 20th week of gestation. Gestational hypertension refers to the new onset of hypertension in the absence of proteinuria, in a previously normotensive pregnant woman at or after 20 weeks of gestation. However, if blood pressure elevation persists by 12 weeks postpartum, the woman is considered to have chronic hypertension. Preeclampsia is defined as gestational blood pressure elevation with proteinuria and usually occurs after 20 weeks of gestation. Proteinuria is defined as the urinary excretion of ≥ 300 mg of protein in a 24-hour period. Preeclampsia superimposed on chronic hypertension is diagnosed when a woman with pre-existing hypertension develops new onset proteinuria after 20 weeks of gestation or after a sudden increase in proteinuria or blood pressure in women with hypertension and proteinuria before 20 weeks’ gestation. Eclampsia refers to the development of grand mal seizures, not attributable to another cause, in a woman with gestational hypertension or preeclampsia. Because blood pressure usually falls during the first half of pregnancy, pre-existing hypertension may not be recognized if the woman is first seen during that time (96). Due to the difficulty in establishing a unique cause for preeclampsia, in spite of many attempts to understand its biologic characteristics and characterize its predictors, heterogeneous causes of gestational hypertensive disorders have been proposed (97). These can be categorized into two major causes: one whose genesis is the result of primarily reduced placental perfusion, and another primarily due to pre-existing maternal disorders, frequently not yet spontaneously evident but unmasked by pregnancy. The placental disorder has a currently unidentified pathobiologic origin. The maternal disorder includes cases with obvious underlying morbidity before pregnancy such as hypertension, renal disease, overweight, diabetes and women with predisposing factors for cardiovascular disease that may not be clinically evident (97). It is not clear whether gestational hypertension and preeclampsia are different diseases with a similar phenotype (hypertension) or if gestational hypertension is an early or mild stage of preeclampsia. However, studies that analysed gestational hypertension and preeclampsia described similar associations with chronic hypertension later in life (98). Studies that analysed preeclampsia and gestational hypertension separately described similar effects, though with stronger associations for preeclampsia (99, 100) which could be interpreted in light of spectrum effects related with the higher severity of this clinical entity. The etiology and pathogenesis of hypertensive disorders in pregnancy is not completely understood. Abnormal trophoblast invasion of uterine blood vessels, immunological intolerance between fetoplacental and maternal tissues, maladaptation 32 to the cardiovascular changes or inflammatory changes of pregnancy, dietary deficiencies, and genetic abnormalities have been proposed as potential etiologies (101, 102). Moreover, the pathophysiologic abnormalities of preeclampsia are numerous, including placental ischemia, generalized vasospasm, abnormal homeostasis with activation of the coagulation system, vascular endothelial dysfunction, abnormal nitric oxide and lipid metabolism, leukocyte activation, and changes in various cytokines as well as in insulin resistance (101, 103). Risk factors for hypertensive disorders in pregnancy represent a selection of antecedents that reflect the disease’s complexity (104). Among the strongest established risk factors for preeclampsia are nulliparity, multifetal gestation and family or personal history of preeclampsia (105, 106). Advanced maternal age is also an established risk factor for adverse outcomes in pregnancy, with all types of hypertensive disorders in pregnancy more frequent in mothers above 35 years old (107). Additionally, family history of hypertension is associated with a doubling risk of preeclampsia (108). The association between excessive weight before and during pregnancy and the risk of hypertensive disorders in pregnancy (100, 109-112) is also already recognized. Despite some contradictory results, other potential risk factors, such as socioeconomic status (113, 114), seasonality (115) and psychosocial stress (116) have been suggested. Worldwide, hypertensive disorders in pregnancy affect approximately 10% of all pregnant women (117, 118). An Australian study (95) reported an overall prevalence of 9.8% pregnancies complicated by hypertension, including 0.6% with chronic hypertension, 4.3% with gestational hypertension, 4.2% with preeclampsia and 0.3% with superimposed preeclampsia on chronic hypertension. Also, a prospective cohort study from the United States of America (109), described a prevalence of gestational hypertension and preeclampsia of 7.5% and 3.6%, respectively. Recently, a study performed in Netherlands (100), restricted to low-risk pregnancies, reported prevalences of 2.1% and 4.1% for preeclampsia and gestational hypertension, respectively. In Portugal, a national survey conducted in 2005 estimated an overall prevalence of hypertensive disorders in pregnancy of approximately 6%, with 1.5% due to chronic hypertension, 2.5% to gestational hypertension, 1.4% to preeclampsia, 0.2% to superimposed preeclampsia and 0.1% to eclampsia (119). Despite the lower rate than described in other countries (95, 100, 109), preterm birth, small for gestational age newborns and foetal death were also more frequent in women with hypertension during pregnancy (119). 33 These major complications of pregnancy increase the risk of adverse obstetric and perinatal outcomes (93-95). In Africa and Asia, nearly one tenth of all maternal deaths are associated with hypertensive disorders of pregnancy, whereas one quarter of maternal deaths in Latin America have been associated with those complications. In European countries, North America, Australia, New Zealand, and Japan hypertensive disorders during pregnancy are the leading cause of maternal death, being responsible for 16.1% of all deaths (120). Considering only Europe, complications of hypertension accounted for 9.2% of maternal deaths, ranging from 2.3% in Germany to 25% in Spain (61). Although there is less information about morbidity, recent data from large national surveillance studies have confirmed that severe obstetric morbidity occurs at a higher rate than maternal mortality (121-124). In fact, all types of hypertensive disorders during pregnancy are associated with severe obstetric morbidity during delivery hospitalizations such as renal failure, pulmonary edema, adult respiratory distress syndrome, puerperal cerebrovascular disorder, disseminated intravascular coagulation syndrome, cardiac arrest or failure, obstetric embolism, sepsis and antepartum or postpartum haemorrhage (94, 95, 125). Likewise, infants exposed to hypertension during pregnancy are more likely to suffer death or major morbidity than those without exposure to hypertension (93, 95). Despite the overall decrease in neonatal mortality during the last decades (93), preeclampsia still carries a 2-fold increased risk of neonatal death and this increased risk has remained relatively constant over time (126). Additionally, offspring of women with hypertension during pregnancy are also at increased risk of preterm birth, being small for gestational age, low Apgar scores, intubation, seizures, respiratory distress syndrome, transient tachypnea, sepsis and neonatal intensive care unit admission (95, 125, 127). A consistent association has been found between preeclampsia and high blood pressure and cardiovascular disease later in women’s life (88, 98). Recently, a systematic review and meta-analysis described that, after preeclampsia, women have an almost fourfold increased risk of hypertension and an approximately twofold increased risk of fatal and nonfatal ischaemic heart disease, stroke, and venous thromboembolism in later life. Furthermore, the overall increase mortality risk after preeclampsia was largely driven by an increased risk of death due to cardiovascular disease (87). Despite some conflicting results, a meta-analysis (128) reported higher systolic and diastolic blood pressure among offspring of women with preeclampsia during pregnancy. Offspring of pregnancies complicated by preeclampsia, between 6 and 19 34 years old, had an increase of 2.3 and 1.7 mmHg in systolic and diastolic blood pressure, respectively, when compared with offspring from normotensive pregnancies. Also, increased risks of hospitalization have been noted in children born to mothers who had preeclampsia (129). A long-term follow-up study showed that people born after pregnancies complicated by gestational hypertension or preeclampsia are at increased risk of stroke, 60 to 70 years after their birth (130). The pathophysiological mechanisms underlying the association between gestational hypertensive disorders and later high blood pressure in women and their children remain not completely understood (64, 131). In women, hypertensive disorders of pregnancy and chronic hypertension share pathways initiated by similar risk factors, such as older age and higher body mass index (106, 132, 133). Additionally, abnormal placentation resulting in reduced perfusion may induce de novo hypertension in the pregnant women, possibly related with an immunologically mediated mechanism for which enhanced exposure to paternal antigens is protective (97). Any of these complications may in turn induce long-term metabolic and vascular abnormalities that increase the overall risk for cardiovascular disease later in life (134). Intrauterine growth restriction, due to mother’s cardiometabolic abnormalities during pregnancy and placental dysfunction, may also increase the risk of vascular dysfunction in the offspring (135-137). In the womb, boys have a different placental growth which puts them at a higher risk of becoming undernourished particularly because their larger size increases nutritional needs. This milieu could increase the vulnerability of boys to consequences of metabolic and vascular derangements of the mother during pregnancy (138). Pregnancy: opportunity for health promotion and disease prevention Pregnancy has been widely referred as a good opportunity for prevention and intervention because of mothers’ strong motivation to protect the wellbeing of the foetus (65, 139). It is well established that adverse lifestyles such as smoking, alcohol intake, poor diet and lack of moderate physical activity are associated with increased risks of adverse maternal and foetal outcomes such as ectopic pregnancy, placental abruption, 35 placenta previa, small for gestational age babies, low birth weight and preterm birth (140-146). The label “teachable moment” has been used to describe naturally occurring life transitions or health events thought to motivate individuals to spontaneously adopt riskreducing health behaviours. McBride et al (65) proposed a model to describe characteristics of effective teachable moments. These are characterized as times that increase perceptions of personal risk and outcome expectancies, prompt strong affective or emotional responses, and redefine self-concept or social roles. In other words, the cognitive response precedes motivation, skills acquisition and self-efficacy that in turn, increase the likelihood of cease the adverse lifestyle. Additional key factors to consider are predisposing factors such as age, dispositional and cultural characteristics that may influence an individual’s cognitive and emotional response (65). Within this model, pregnancy is conceptualized as a powerful “teachable moment”. Pregnancy provides an immediate and personal experience of risk that is related with the health of the mother and baby, which enhances the perceived need of adopting healthy lifestyles (147). According to a study performed in the United Kingdom between 1998 and 2003 (148), there was a significant reduction in smoking, alcohol consumption and intake of caffeinated drinks when women became pregnant, although little change occurred in fruit and vegetable intake. In fact, before pregnancy, 27% of women smoked, 54% of women drank more than 4 units of alcohol per week and 39% had estimated intakes of caffeine in drinks of more than 300mg per day, whereas comparable figures for early pregnancy were 15%, 10% and 16% respectively. The emotional responses that surround pregnancy may also make it an opportune time to initiate change. Emotional responses are thought to influence an individual’s judgment about the significance and meaning of an event (149).Therefore, pregnancy may prompt feelings of concern about the well-being of the foetus, which may motivate women to change their lifestyles. Finally, pregnancy is a time when personal and social roles change as women become mothers in addition to their other roles. Women are adopting the maternal role, which carries expectations for major changes in lifestyle and self-image (147). Worldwide, smoking is the single most frequent preventable cause of death (150) and smoking cessation reduces the long-term risk of cancer and cardiovascular diseases (151, 152). Cigarette smoking during pregnancy is associated with increased risks for both the mother and the newborn (140-142). Therefore, mothers tend to be 36 strongly motivated to protect the health of the unborn baby and there is important social pressure to quit smoking during pregnancy (66, 139, 153). The EURO-PERISTAT study, that monitored and evaluated perinatal health in Europe, reported prepregnancy smoking prevalence levels in 2004 ranging from 7.9% in Lithuania to 35.9% in France (61). In Portugal, according to a study performed in 25 public maternity units, 30% of women smoked before pregnancy in Portugal (154). In many countries in Europe, more than 10% of women smoke during pregnancy (61). In Portugal, 14.7% of women smoked during the third trimester of pregnancy in 2004 (61). Previous literature reported smoking cessation rates ranging from 30% to 50% in different countries (155-158). Predictors of smoking cessation among pregnant women have been well investigated, internationally. Women who quit smoking during pregnancy are more likely to report being married or in a stable relationship and having higher educational qualifications (156, 157, 159). However, a study that assessed the reporting of prenatal smoking in birth certificates and in confidential questionnaires found that more educated women are less willing to admit smoking to prenatal care providers, which may lead to overestimation of the association between education and smoking cessation during pregnancy (160). Quitting in pregnancy appears to be more successful in first pregnancies regardless of social background, with the probability decreasing with the number of pregnancies (161). It has been suggested that past pregnancies resulting in the birth of a healthy child despite tobacco consumption might weaken the women’s motivation to change their smoking habits in subsequent pregnancies (162). Also, women receiving adequate prenatal care show a higher rate of smoking cessation (163, 164). Despite the impact of adequate prenatal care on smoking cessation, this may also be due to personal characteristics of women since those more prone to quit smoking are also probably more likely to attend prenatal care (165). Those who initiate smoking earlier or were heavier smokers were more likely to smoke during pregnancy, which may emphasize the physiological and psychological role of nicotine dependence in smoking cessation (166). Additionally, concern about weight gain during and after pregnancy may be a factor that interferes with smokers’ efforts to quit during pregnancy (167), with pregnant smokers reporting the use of smoking as a weight management strategy (168). Smoking cessation during pregnancy could be a unique opportunity in women life to achieve long-term sustainability of smoking cessation, with clear benefits for their future health. However, despite a high proportion of women quitting their tobacco consumption during pregnancy the rate of relapse or return to previous rates following childbirth is high. In fact, approximately 25% of women who quit smoking during 37 pregnancy will relapse within one month of delivery (157). By three months postpartum, 40% to 50% of women will have returned to smoke and, 70% will have experienced a relapse within one year postpartum (156, 157, 169). It has been previously described that younger women, divorced or without a partner and with a lower education were more likely to resume smoking (156, 157). Furthermore, breastfeeding may help to prevent or delay postpartum smoking relapse (158), by educating women about the benefits of breastfeeding and by facilitating and encouraging continued breastfeeding throughout the first postpartum year (170). The health of the offspring can also constitute a motivation to stop smoking. A substantial body of evidence supports that involuntary tobacco smoke exposure adversely affects children’s respiratory health by increasing the risk of respiratory infections (171, 172). In uterus exposure to maternal smoking was associated with an almost two-fold increased prevalence of physician-diagnosed asthma (171), while children exposed after birth had more often rhinitis than children of non-smoking parents (172). In Portugal, smoking remains more frequent among men but its prevalence is increasing in women (173). The prevalence is higher among younger and more educated women placing Portugal at a relatively early stage of the epidemic compared to other Western European countries (174, 175). Therefore, it is expected that, in the near future, the burden of tobacco-related illness will increase substantially (176). In order to guide the development of preventive strategies to reduce the future burden of smoking and promote the health of children and families, it is important not only to describe exposure to smoking in pregnancy and postpartum but also to identify modifiable determinants of such behaviours. 44 45 This PhD thesis is based on the cohort Geração XXI, the first prospective Portuguese population-based birth cohort. A general description of the participants and methods in this cohort study is provided below. The selection of participants eligible for each analysis depended on the specific objectives of the investigations and is described in detail in the methods sections of the individual papers. A SSEMBLING OF THE BIRTH COHORT The birth cohort Geração XXI was assembled between 2005 and 2006 at all 5 public maternity units covering the metropolitan area of Porto, Portugal: Centro Hospitalar de Vila Nova de Gaia (CHVNG), Centro Hospitalar do Porto - Maternidade de Júlio Dinis (MJD), Hospital de São João (HSJ), Centro Hospitalar do Porto - Hospital de Santo António (HSA), and Unidade Local de Saúde de Matosinhos – Hospital Pedro Hispano (HPH). The maternities corresponded to level III units, with differentiated perinatal support, and all except MJD, were included in a general hospital, with a variety of medical and surgical specialties. Trained interviewers at the five hospitals were responsible for the study presentation and subsequent invitation of the mothers after delivery. Fathers were also invited to participate. Of the invited mothers 91.4% accepted to participate. A total of 8647 infants, corresponding to 8495 mothers, were enrolled in the cohort. D ATA COLLECTION Mothers were invited to participate after delivery, during the hospital stay. Data on demographic and socioeconomic characteristics (age, marital status, years of formal education, occupation), personal and family medical history, gynaecologic and obstetric history (age at menarche, number of previous pregnancies and births), prenatal care (number and place of the prenatal care, medication, and symptoms and pregnancyrelated co-morbidities), and lifestyles (including smoking, alcohol intake and illicit drug 46 use) were collected within 72 hours after delivery, in a face-to-face interview, conducted by trained interviewers using structured questionnaires. Clinical records were also reviewed at birth by the same interviewers to retrieve data on prenatal care, pregnancy complications (gestational hypertensive disorders, gestational diabetes and placental abnormalities), delivery (type of delivery and medication) and neonatal characteristics (birth weight and gestational age). Fathers were also invited to answer a self-administered questionnaire focused on demographic and social conditions, lifestyles and medical history. Anthropometrics were performed, by the same interviewers, to the mother, the father and their newborn. After delivery, blood samples were collected from the mother and the umbilical cord. Fathers were also requested to provide a venous blood sample. Plasma and serum samples were distributed in aliquots and a whole blood sample was maintained for each member of the family. All blood samples were put into transitory storage at - 20ºC and later stored at -80ºC. M ISSING DATA RECOVERY Between October 2008 and June 2009, missing data on the questionnaires were recovered through the review of obstetrical records (PAPER I). We reviewed the medical records of 3657 women with at least one missing value on personal history of disease, anthropometrics, pregnancy complications, blood glucose or oral glucose tolerance test results during pregnancy, or newborn characteristics at birth, and assessed the agreement of information reported by participants with data from medical records. All the clinical records were reviewed by one trained abstractor, who had been an interviewer at the baseline evaluation of Geração XXI, using standardized criteria. Inter-observer variability was assessed using 400 clinical records randomly selected among the 3657 reviewed. Data were collected independently by 2 trained abstractors, both interviewers at birth. Data on pregnancy pathological complications and mothers’ anthropometrics were successfully recovered. Agreement between questionnaire and records in family history data was fair, particularly for cardiovascular disease [k=0.27; 95%confidence interval (95%CI) 0.23-0.32]. Personal history of diabetes was highly concordant (k=0.82; 95%CI 0.70-0.93), while hypertension was moderately concordant (k=0.60; 47 95%CI 0.50-0.69). In general, the discrepancies found for weight and height indicated higher values in the clinical record than in the questionnaire and led to discrepancies in prepregnancy body mass index classes of 10.3% of women (weighted k=0.82; 95%CI 0.80-0.83). Data were very consistent between reviewers, with highest agreement for gestational diabetes (k=1.00) and birth weight (99.5% concordant). The differences observed for women’s weight and height were negligible, not affecting the global classification by body mass index categories before pregnancy (weighted k=0.99; 95%CI 0.98-1.00). R E - EVALUATION OF THE BIRTH COHORT AT 4 YEARS Between April 2009 and April 2011, an average of 4 years after birth, all the children with their mothers were invited to attend the re-evaluation of the cohort. During this evaluation period, participants were invited to an interview and physical examination. If they were unable to participate, they were invited to answer a telephone interview, covering part of the self-reported data collected in the default interview. During the follow-up period, 5986 (69.2%) of the children attended a face faceto-face interview, 1472 (17.0%) provided self-reported data by telephone interview, and 1189 (13.7%) were lost to follow-up. Regarding the mothers, 5729 (67.4%) participated in the face face-to-face interview, 1428 (16.8%) provided self-reported data by telephone interview and 1338 (15.8%) were lost to follow-up. Figure 4 presents the maternal participation at follow-up, according to the different strategies implemented. Overall, 86.2% of the children and 84.2% of the mothers were re-evaluated. The comparison of socio-demographic characteristics between participant and non-participant mothers at follow-up and among mothers who attended a face-to-face interview and who answered a telephone interview are presented in Table 1. Mothers who attended the follow-up re-evaluation were older, more likely to be pregnant for the first time at baseline and married, with a higher educational level, and more frequently employed, than those who did not participate in the re-evaluation. When comparing mothers evaluated by the two strategies of data collection, those who answered a faceto-face interview were older, more educated and more frequently employed. 48 Figure 4. Maternal participation at follow-up, according to the different strategies implemented. Table 1. Comparison of socio-demographic characteristics between participant and nonparticipant mothers at follow-up and, among the former, between mothers who attended a faceto-face interview and who answered a telephone interview . Non participants Fac e - to - face interview Telephone interview n(%) n(%) n(%) p* p** Age (years), n (%) <0.001 <0.001 < 25 487 (36.7) 1010 (17.7) 356 (25.0) 25-29 368 (27.7) 1723 (30.2) 441 (31.0) 30-34 322 (24.3) 1961 (34.3) 406 (28.5) ≥ 35 150 (11.3) 1019 (17.8) 220 (15.5) Gravidity , n (%) <0.001 0.205 1 607 (45.7) 2778 (48.7) 679 (47.8) 2 423 (31.9) 1969 (34.5) 475 (33.4) ≥ 3 298 (22.4) 963 (16.9) 268 (18.9) Marital status, n (%) <0.001 0.058 Married/living with a partner 1196 (90.4) 5389 (94.5) 1320 (93.2) Single/ divorced/widow 127 (9.6) 312 (5.5) 96 (6.8) Education (years), n (%) <0.001 <0.001 ≤ 4 159 (12.0) 371 (6.5) 133 (9.4) 5-9 672 (50.9) 2181 (38.4) 634 (44.8) 10-12 312 (23.6) 1571 (27.7) 370 (26.1) > 12 178 (13.5) 1551 (27.3) 279 (19.7) Working condition , n (%) <0.001 <0.001 Employed 760 (57.4) 4238 (74.9) 989 (70.0) Unemployed 382 (28.9) 1017 (18.0) 276 (19.5) Housewife 122 (9.2) 268 (4.7) 103 (7.3) Others 60 (4.5) 132 (2.3) 45 (3.2) * Participation vs. non participation ** Face-to-face interview vs. telephone interview Note: In each variable, the total may not add to 8495 due to missing dat a 49 D ATA COLLECTION In the face-to-face interview, information was collected by trained interviewers, using two structured questionnaires: one concerning the child’s health and another regarding the health of the mother. The mothers’ questionnaire comprised data on demographic characteristics (marital status and owning one’s house), personal and family medical history, gynaecologic and obstetric history (number of pregnancies and births before and after the index pregnancy), and lifestyles (including smoking, alcohol intake, physical activity and diet). The children’s questionnaire included data on socioeconomic characteristics of the parents (years of formal education, occupation) and on child’s development, health and habits (including sleeping hours, physical activity and diet). Mothers and children also underwent a physical examination, including anthropometric evaluation, blood pressure measurement and a venous blood sample withdrawal. Children were submitted, additionally, to tetrapolar bioimpedance analysis. The venous blood sample was drawn after 12-hour overnight fast and all the samples were analyzed at the central laboratory of Hospital de São João. The telephone interview also contained data regarding the health of the mother and the children, although with a restricted number of questions. Information was collected by trained interviewers, who also performed face-to-face interviews, using a structured questionnaire. Data on medical and obstetric history (pregnancies and births after the index pregnancy) as well as smoking status and self-reported weight and height were collected for the mothers. Regarding the child, data on socioeconomic characteristics of the parents (years of formal education, occupation), and on child’s development, health and habits (including sleeping hours, physical activity and diet) were collected. Q UALITY CONTROL In both evaluation periods, all interviewers were rigorously trained using a structured protocol addressing all the questionnaires’ queries and periodic supervision of their work was undertaken. At baseline, interviewers were responsible for the application of the questionnaires to the mothers, the distribution of self-administered 50 questionnaires to the fathers and the anthropometric evaluations of the parents and newborns, having received specific and intensive training for these tasks. Interviewers were also responsible for managing the blood samples, namely their coded identification and storage. During follow-up reevaluation, interviewers were responsible for the application of the questionnaires to the mothers and for the physical examination both of the mothers and their children. A multidisciplinary team, including physicians, psychologists, nutritionists and pharmacists, with experience in other national and international projects, were responsible for the staff training and the development of the questionnaires applied at baseline and at follow-up. S TATISTICAL ANALYSES All statistical analyses were performed using Stata 9.0 (College Station, TX, 2005). Overall, we used classical inferential statistics to answer the objectives of this PhD thesis. Data were described as counts and proportions for categorical variables, mean and standard deviation (SD) for normally distributed continuous variables, and median and interquartile range (IQR) for non-normally distributed continuous variables. The prevalence of the outcomes is presented with 95% confidence intervals (95% CI), as well as the incidence rates of hypertension, which were reported per 1000 personyears. According to the specific objectives established, different analytic approaches were used and are described in detail in the methods sections of the papers. The prevalence ratio (PR) is more conservative, consistent, and interpretable than the odds ratio (OR) when prevalences are to be compared between groups (177, 178). For this reason, we privileged the estimate of PR in our models. Therefore, in Papers III and V crude and adjusted prevalence ratios (PR) were computed, using robust Poisson regression. However, PR applies only to dichotomous outcomes, whereas categorical dependent variables with more than two classes request multinomial logistic regression models. Thus, in Papers II and VI, to privilege homogeneity in the analysis among all the outcomes, we used logistic regression to estimate OR for all the risk factors 51 considered, binary for dichotomous outcomes and multinomial for 3-class outcomes. In order to assess the potential for prevention, population attributable fractions were calculated using the command punaf (179) in Stata, based on the adjusted PR and putative alternative distributions of exposure (Paper III). In Paper IV, to estimate adjusted mean differences in systolic and diastolic blood pressure, for mothers and offspring, according to gestational hypertensive disorders, multiple linear regression was used. The association between gestational hypertensive disorders and the mother’s incidence of hypertension was estimated by crude and adjusted incidence rate ratios (IRR) and respective 95% CI, using Poisson regression. E THICAL CONSIDERATIONS The study protocols for both evaluation periods were approved by the Ethics Committee of Hospital de São João and by the Portuguese Authority of Data Protection. Procedures were developed in order to guarantee data confidentiality and protection. All participants received an explanation on the purposes and design of the study, and gave written informed consent at baseline and at follow-up evaluation. The children gave it through their legal representative. Verbal consent was explicitly solicited at the beginning of telephone interviews. 52 53 PAPER I Alves E, Lunet N, Correia S, Morais V, Azevedo A, Barros H. Inter-rater variability in medical record review and agreement with self-reported data collected by questionnaire. Gac Sanit 2011; 25(3):211-9 60 61 62 63 64 65 RESULTS 66 67 PAPER II Alves E, Correia S, Barros H, Azevedo A. Prevalence of self-reported cardiovascular risk factors in Portuguese women: a survey after delivery. Int J Public Health 2012 [Epub ahead of print] 68 69 76 77 78 79 80 81 PAPER III Alves E, Azevedo A, Rodrigues T, Santos AC, Barros H. Impact of risk factors on hypertensive disorders in pregnancy, in primiparae and multiparae. [submitted] 82 83 Impact of risk factors on hypertensive disorders in pregnancy, in primiparae and multiparae Elisabete Alves 1,2 , Ana Azevedo 1,2 , Teresa Rodrigues 1,2,3 , Ana Cristina Santos 1,2 , Henrique Barros 1,2 1 Department of Clinical Epidemiology, Predictive Medicine and Public Health, University of Porto Medical School, Porto, Portugal 2 Institute of Public Health – University of Porto (ISPUP), Porto, Portugal; 3 Department of Obstetrics and Gynecology, Centro Hospitalar São João EPE, Porto, Portugal; 84 ABSTRACT Purpose: To assess the impact of age, education, family history of cardiovascular disease, prepregnancy overweight/obesity and weight gain during pregnancy on hypertensive disorders, among primiparae and multiparae. Methods: In a birth cohort study, puerperae were consecutively recruited at all public maternities of Porto, Portugal (2005-2006). We included 6952 women with singletons and complete data on key variables. Hypertensive disorders included chronic hypertension, gestational hypertension or preeclampsia/eclampsia. Prevalence ratios were computed using Poisson regression and population attributable fractions were calculated. Results: Overall, hypertensive disorders affected 4.6% of single pregnancies, and were associated with older age, lower education, family history of cardiovascular disease and excessive weight before and during pregnancy, similarly in primiparae and multiparae. Approximately 50% of cases among primiparae and 70% among multiparae were attributable to the joint effect of pregnancies after 34 years of age, education below 12 years, family history of cardiovascular disease and excessive weight before and during pregnancy. Conclusions: The risk factors explained a high proportion of hypertensive disorders during pregnancy. Excessive weight before and during pregnancy had a very large contribution. The substantial joint effect of the risk factors suggests that interventions focusing on these risk factors should be part of pre-conceptional and prenatal care. Keywords: Cohort studies; Hypertension, Pregnancy-Induced; Risk factors 85 INTRODUCTION Hypertensive disorders complicating pregnancy include chronic and gestational hypertension, preeclampsia and eclampsia, with the latter three entities representing generally transient hypertensive conditions with onset during pregnancy (1). These major complications of pregnancy increase the risk of adverse obstetric and perinatal outcomes (2, 3, 4) and are associated with high blood pressure and cardiovascular disease later in women’s life (5, 6), with approximately half of those with preeclampsia developing chronic hypertension (7). In Portugal, a national survey conducted in 2005 estimated an overall prevalence of hypertensive disorders in pregnancy of approximately 6% (8). Although this rate was lower than in other countries (4, 9, 10), preterm birth, small for gestational age newborns and fetal death were more frequent in women with hypertensive pregnancies (8). Maternal socio-demographic and metabolic factors, as well as family history of hypertension, have been reported as determinants of hypertension in pregnancy, but the etiology of hypertensive disorders of pregnancy is not completely understood (11, 12, 13). Primiparity is consistently associated with preeclampsia (12) while other risk factors, usually related with chronic hypertension in the general population, could be expected to have a stronger effect in multiparae. Also, the interdependence of these risk factors, resulting in their clustering (14), raises the question of the added predictive value of considering several factors simultaneously. We are unaware of studies considering the joint impact of several risk factors on the risk of hypertension during pregnancy. Thus, we assessed the impact of age, education, family history of cardiovascular disease, prepregnancy BMI and pregnancy weight gain on hypertensive disorders in pregnancy, in primiparous and multiparous mothers of a Portuguese birth cohort. In order to assess the potential for prevention, we estimated the population 92 DISCUSSION In this study of Portuguese puerperae, 5% of primiparae and 4% of multiparae had a hypertensive disorder in pregnancy. Approximately 50% of cases among primiparae and 70% among multiparae were attributable to the joint effect of pregnancies after 34 years of age, low education, family history of cardiovascular disease, prepregnancy overweight/obesity and excessive weight gain during pregnancy. The associations between the risk factors and hypertensive disorders during pregnancy were not different between primiparae and multiparae, but overall the impact was higher among multiparous reflecting the higher prevalence of risk factors among those women. This is the first study that approaches the impact of several risk factors on hypertensive disorders in pregnancy in a large population-based cohort study. However, some limitations should be pointed. In our study, the prevalence of hypertensive disorders during pregnancy was lower than reported in other industrialized nations (4, 9, 10), and also slightly lower than the 5.6% previously described in a national survey for the Portuguese population (8) Since the authors reported no significant differences in prevalence across the country (8) we expected to find a similar prevalence. Despite the high proportion of participation in our cohort, it is possible that the refusals occurred mainly in cases with higher maternal and/or perinatal complications in the post-partum. Since hypertensive disorders are associated with increased risks of adverse outcomes (2, 3, 4) their prevalence may have been underestimated. The absence of objective measurements of blood pressure could also lead to this underestimation. However, 99.3% of the women had more than three prenatal visits and 85% attended the first appointment before 12 weeks of gestation (results not shown); thus, it is unlikely that a diagnosis of hypertension was missed. Additionally, considering a combined endpoint of hypertensive disorders during pregnancy is an advantage, because even if some cases of chronic hypertension were 93 misclassified as gestational, due to the low awareness of hypertension in young ages (20) they were still identified as a hypertensive disorder in our sample. The higher proportion of missing data among those with lower education, which is also associated with the outcome, raises the possibility of a selection bias. In this case, we may be underestimating the prevalence and, consequently, the association of hypertensive disorders in pregnancy among less educated women. Also, since prepregnancy weight was self-reported after delivery and weight tends to be underreported by women (21), the prevalence of overweight and obesity may be underestimated; if this information bias affected hypertensive women less than nonhypertensives, the differential effect could have contributed to an overestimation of the association. However, it is unlikely that such effect would explain an almost 4-fold increase in prevalence. We observed a positive association between delivering a male newborn and hypertensive disorders during pregnancy among primiparae. Heterogeneous causes of gestational hypertensive disorders have been proposed, comprising both a background increased vascular risk possibly with a yet unidentified higher blood pressure beforehand and hypertension induced by disturbed perfusion of the placenta (22). In light of this causal model, the role of an immune response to paternal antigens in a de novo and transient hypertensive state could be expected to occur more often with male babies. The lack of association in multiparae may be explained by the sex of previous children, which was not assessed. In contrast, the associations between all other risk factors and hypertensive disorders during pregnancy were not different between primiparae and multiparae. Although most previous research has focused on preeclampsia (23), the similarities in risk factors according to parity support that gestational hypertension and preeclampsia/eclampsia represent different stages of one pathological process, as previously proposed (24). Studies that analyzed these two entities separately described similar effects, though with stronger associations for preeclampsia (10, 25) which could 94 be interpreted in light of spectrum effects related with the higher severity of this clinical entity. When analyzing chronic hypertension as the sole outcome, we found no significant differences regarding the associations with the risk factors considered in comparison with their association with gestational hypertension and preeclampsia/eclampsia (data not shown). It could be hypothesized that, even if a specific pregnancy-induced hypertension due to disturbed circulation in the placenta exists, as discussed above, the risk would be higher in women with more classical risk factors for high blood pressure. Overall, 30.6% and 36.5% of hypertensive cases were attributable to overweight or obesity, among primiparae and multiparae, respectively. Our study confirms the association between excessive weight before and during pregnancy and the risk of hypertensive disorders in pregnancy (9, 10, 23, 26, 27). In a Danish study, overweight independently contributed 9.2% and obesity 11.0% for the occurrence of preeclampsia in primiparae (26). Similarly, among multiparae, overweight and obesity were responsible for 8.3% and 10.9% cases of preeclampsia, respectively. In a population of Latin women, those who gained more weight than recommended had a 3fold increased risk of gestational hypertension and a 4-fold increased risk of preeclampsia (27). In our obstetric population, overweight, obesity and weight gain during pregnancy above recommended were common and had a considerable independent impact on hypertension in pregnancy. These metabolic conditions constitute main modifiable risk factors at the individual level, with a vast potential for prevention before and during pregnancy. Information regarding lifestyles, such as physical exercise and diet were not available in our cohort. However, BMI is a good surrogate for these lifestyles and is much more objective and easy to quantify reliably. Socio-demographic and metabolic factors, as well as family history of cardiovascular disease were positively associated with hypertensive disorders during pregnancy in both primiparae and multiparae. Maternal age is an established risk factor for adverse outcomes in pregnancy, with all types of hypertensive disorders more frequent in 95 mothers above 35 years old (28). Family history of hypertension is associated with a doubling risk of preeclampsia (29). Similarly, a first-degree family history of vascular risk increases the likelihood of developing preeclampsia and gestational hypertension (30). Despite some contradictory results, probably due to differences in exposure or outcome definition, prospective data from a large Dutch population-based cohort showed that women with low educational level were more likely to develop gestational hypertension (31) and preeclampsia (32) than women with higher educational level. The association between education and pregnancy hypertension may represent an early manifestation of the socioeconomic differences in cardiovascular morbidity and mortality in women (33). Despite similar associations, the impact of low education and family history of cardiovascular disease was approximately double in multiparous, while the effect of age increased from 5 to 29%, due to the higher proportion of pregnancies after 35 years of age in multiparae. We recognize that, regarding these risk factors, the potential for change is mainly theoretical or only effective at the global community level. However, a stricter control of overall risk could be attained if health education was sensitive to literacy (34). In conclusion, the five risk factors explained a high proportion of hypertensive disorders during pregnancy. Despite the relatively low prevalence of these conditions, the severity of shortand long-term consequences for both mother and child emphasizes the importance of control of this outcome. Excessive weight before and during pregnancy, the most modifiable of all factors considered, had a very large contribution, particularly among primiparae. The substantial joint effect of the risk factors emphasizes the need to improve patient education and suggests that interventions focusing on these risk factors should be part of pre-conceptional and prenatal care in order to prevent maternal and perinatal complications. 96 REFERENCES 1. Report of the National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy. Am J Obstet Gynecol 2000; 183(1):S1-S22. 2. Ananth CV, Basso O. Impact of pregnancy-induced hypertension on stillbirth and neonatal mortality. Epidemiology 2009; 21(1):118-123. 3. Kuklina EV, Ayala C, Callaghan WM. Hypertensive disorders and severe obstetric morbidity in the United States. Obstet Gynecol 2009; 113(6):1299-1306. 4. Roberts CL, Algert CS, Morris JM, Ford JB, Henderson-Smart DJ. Hypertensive disorders in pregnancy: a population-based study. Med J Aust 2005; 182(7):332-335. 5. Magnussen EB, Vatten LJ, Smith GD, Romundstad PR. Hypertensive disorders in pregnancy and subsequently measured cardiovascular risk factors. Obstet Gynecol 2009; 114(5):961-970. 6. Wilson BJ, Watson MS, Prescott GJ, et al. Hypertensive diseases of pregnancy and risk of hypertension and stroke in later life: results from cohort study. BMJ 2003; 326(7394):845. 7. Bellamy L, Casas JP, Hingorani AD, Williams DJ. Pre-eclampsia and risk of cardiovascular disease and cancer in later life: systematic review and meta-analysis. BMJ 2007; 335(7627):974. 8. Povoa AM, Costa F, Rodrigues T, Patricio B, Cardoso F. Prevalence of hypertension during pregnancy in Portugal. Hypertens Pregnancy 2008; 27(3):279-284. 9. Bodnar LM, Catov JM, Klebanoff MA, Ness RB, Roberts JM. Prepregnancy body mass index and the occurrence of severe hypertensive disorders of pregnancy. Epidemiology 2007; 18(2):234-239. 10. Gaillard R, Steegers EA, Hofman A, Jaddoe VW. Associations of maternal obesity with blood pressure and the risks of gestational hypertensive disorders. The Generation R Study. J Hypertens 2011; 29(5):937-944. 97 11. Zhang J, Zeisler J, Hatch MC, Berkowitz G. Epidemiology of pregnancy-induced hypertension. Epidemiol Rev 1997; 19(2):218-232. 12. Duckitt K, Harrington D. Risk factors for pre-eclampsia at antenatal booking: systematic review of controlled studies. BMJ 2005; 330(7491):565. 13. Wolf M, Shah A, Jimenez-Kimble R, Sauk J, Ecker JL, Thadhani R. Differential risk of hypertensive disorders of pregnancy among Hispanic women. J Am Soc Nephrol 2004; 15(5):1330-1338. 14. Lloyd-Jones DM, Leip EP, Larson MG, et al. Prediction of lifetime risk for cardiovascular disease by risk factor burden at 50 years of age. Circulation 2006; 113(6):791-798. 15. Pinto E, Ramos E, Severo M, et al. Measurement of dietary intake of fatty acids in pregnant women: comparison of self-reported intakes with adipose tissue levels. Ann Epidemiol 2010; 20(8):599-603. 16. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults: executive summary. Expert Panel on the Identification, Evaluation, and Treatment of Overweight in Adults. Am J Clin Nutr 1998; 68(4):899917. 17. Rasmussen KM, Yaktine AL, eds. Weight Gain During Pregnancy: Reexamining the Guidelines. Ed. Washington (DC): National Academies Press (US); 2009. 18. Alves E, Lunet N, Correia S, Morais V, Azevedo A, Barros H. Medical record review to recover missing data in a Portuguese birth cohort: agreement with selfreported data collected by questionnaire and inter-rater variability. Gac Sanit 2011; 25(3):211-219. 19. Newson R. Population attributable and unattributable fractions for cohort and cross-sectional studies. Available at http://fmwww.bc.edu/repec/bocode/p/punaf.html. Retrieved June 16, 2011. 98 20. Macedo ME, Lima MJ, Silva AO, Alcantara P, Ramalhinho V, Carmona J. Prevalence, awareness, treatment and control of hypertension in Portugal: the PAP study. J Hypertens 2005; 23(9):1661-1666. 21. Brunner Huber LR. Validity of self-reported height and weight in women of reproductive age. Matern Child Health J 2007; 11(2):137-144. 22. Ness RB, Roberts JM. Heterogeneous causes constituting the single syndrome of preeclampsia: a hypothesis and its implications. Am J Obstet Gynecol. 1996; 175(5):1365-1370. 23. O'Brien TE, Ray JG, Chan WS. Maternal body mass index and the risk of preeclampsia: a systematic overview. Epidemiology 2003; 14(3):368-374. 24. Villar J, Carroli G, Wojdyla D, et al. Preeclampsia, gestational hypertension and intrauterine growth restriction, related or independent conditions? Am J Obstet Gynecol 2006; 194(4):921-931. 25. Bhattacharya S, Campbell DM, Liston WA. Effect of Body Mass Index on pregnancy outcomes in nulliparous women delivering singleton babies. BMC Public Health. 2007; 7:168. 26. Catov JM, Ness RB, Kip KE, Olsen J. Risk of early or severe pre-eclampsia related to pre-existing conditions. Int J Epidemiol 2007;36(2):412-419. 27. Fortner RT, Pekow P, Solomon CG, Markenson G, Chasan-Taber L. Prepregnancy body mass index, gestational weight gain, and risk of hypertensive pregnancy among Latina women. Am J Obstet Gynecol 2009; 200(2):167 e161-167. 28. Joseph KS, Allen AC, Dodds L, Turner LA, Scott H, Liston R. The perinatal effects of delayed childbearing. Obstet Gynecol 2005; 105(6):1410-1418. 29. Magnussen EB, Vatten LJ, Lund-Nilsen TI, Salvesen KA, Davey Smith G, Romundstad PR. Prepregnancy cardiovascular risk factors as predictors of preeclampsia: population based cohort study. BMJ 2007; 335(7627):978. 99 30. Ness RB, Markovic N, Bass D, Harger G, Roberts JM. Family history of hypertension, heart disease, and stroke among women who develop hypertension in pregnancy. Obstet Gynecol 2003; 102(6):1366-1371. 31. Silva L, Coolman M, Steegers E, et al. Maternal educational level and risk of gestational hypertension: the Generation R Study. J Hum Hypertens 2008; 22(7):483492. 32. Silva LM, Coolman M, Steegers EA, et al. Low socioeconomic status is a risk factor for preeclampsia: the Generation R Study. J Hypertens 2008; 26(6):1200-1208. 33. Mackenbach JP, Cavelaars AE, Kunst AE, Groenhof F. Socioeconomic inequalities in cardiovascular disease mortality; an international study. Eur Heart J 2000; 21(14):1141-1151. 34. Osborn CY, Paasche-Orlow MK, Bailey SC, Wolf MS. The mechanisms linking health literacy to behavior and health status. Am J Health Behav 2011; 35(1):118-128. 100 Table 1. Characteristics of the participants, according to parity. Parity Overall Primiparae Multiparae n (%) n (%) n (%) p Age (years) <25 1504 (21.6) 1164 (29.6) 340 (11.3) 25-29 2119 (30.5) 1423 (36.2) 696 (23.0) 30-34 2210 (31.8) 1059 (27.0) 1151 (38.1) ≥35 1119 (16.1) 283 (7.2) 836 (27.6) <0.001 Marital status Married/cohabiting 6512 (94.0) 3575 (91.3) 2937 (97.5) Single/divorced/widow 416 (6.0) 339 (8.7) 77 (2.6) <0.001 Education (years) <5 493 (7.1) 110 (2.8) 383 (12.7) 5-9 2883 (41.5) 1491 (38.0) 1392 (46.0) 10-11 528 (7.6) 337 (8.6) 191 (6.3) ≥12 3048 (43.8) 1991 (50.7) 1057 (35.0) <0.001 Employment status Employed 4973 (71.7) 2904 (74.1) 2069 (68.6) Unemployed 1374 (19.8) 737 (18.8) 3637 (21.1) Housewife 385 (5.6) 104 (2.6) 281 (9.3) Other (student/retired) 205 (3.0) 175 (4.5) 30 (1.0) <0.001 Income ( €/month) <500 435 (6.4) 210 (5.5) 225 (7.6) 500-1000 2002 (29.4) 1086 (28.2) 916 (30.8) 1001-1500 1767 (25.9) 1052 (27.4) 715 (24.1) ≥1501 1943 (28.5) 1101 (28.6) 842 (28.3) Does not know/ Prefers not to answer 669 (9.8) 395 (10.3) 274 (9.2) <0.001 Family history of cardiovascular disease § No 5820 (83.7) 3408 (86.7) 2412 (79.8) Yes 1132 (16.3) 7521 (13.3) 611 (20.2) 0.001 Newborn’s sex Male 3548 (51.0) 2030 (51.7) 1518 (50.2) Female 3404 (49.0) 1899 (48.3) 1505 (49.8) 0.230 Smoking status Never smoker 4303 (61.9) 2433 (61.9) 1870 (61.9) Current smoker 1855 (26.7) 1066 (27.1) 789 (26.1) Ex-smoker 794 (11.4) 430 (10.9) 364 (12.0) 0.289 Prepregnancy BMI (Kg/m 2 ) <25.0 4867 (70.0) 2943 (74.9) 1924 (63.6) 25.0-29.9 1475 (21.2) 717 (18.2) 758 (25.1) ≥30 610 (8.8) 269 (6.9) 341 (11.3) <0.001 Weight gain during pregnancy ¥ As recommended 2536 (36.5) 1425 (36.3) 1111 (36.8) Bellow recommended 1796 (25.8) 916 (23.3) 880 (29.1) Above recommended 2620 (37.7) 1588 (40.4) 1032 (34.1) <0.001 BMI, body mass index § Reporting at least one parent or sibling affected by stroke and/or myocardial infarction ¥ According to the Institute of Medicine recommendations (2009) Note: In each variable, the total may not add to 6952 due to missing data 101 Table 2. Crude and adjusted prevalence ratios for the association between baseline characteristics of women, pregnancy and newborn, and pregnancy complicated by hypertension, among primiparous and multiparous women. 95%CI, 95% confidence interval; BMI, body mass index; PR, prevalence ratio * Adjusted for all the variables in the table ¥ According to the Institute of Medicine recommendations (2009) Primiparae Multiparae n (%) Crude PR (95%CI) Adjusted PR* (95%CI) n (%) Crude PR (95%CI) Adjusted PR* (95%CI) p for interaction Age (years) < <25 45 (3.9) 1 1 6 (1.8) 1 1 25-29 73 (5.1) 1.33 (0.92-1.91) 1.23 (0.83-1.82) 21 (3.0) 1.71 (0.70-4.20) 1.51 (0.61-3.78) 30-34 55 (5.2) 1.34 (0.91-1.97) 1.25 (0.82-1.91) 43 (3.7) 2.12 (0.91-4.93) 1.92 (0.81-4.59) ≥ 35 22 (7.8) 2.01 (1.23-3.29) 1.82 (1.07-3.10) 55 (6.6) 3.73 (1.62-8.58) 2.98 (1.26-7.08) 0.268 Education (years) ≥12 87 (4.4) 1 1 33 (3.1) 1 1 10-11 15 (4.6) 1.02 (0.60-1.74) 1.05 (0.60-1.83) 7 (3.7) 1.17 (0.53-2.62) 1.15 (0.50-2.62) 5-9 80 (5.4) 1.23 (0.91-1.65) 1.22 (0.88-1.70) 59 (4.2) 1.36 (0.89-2.06) 1.31 (0.85-2.04) <5 13 (11.8) 2.70 (1.56-4.69) 2.04 (1.12-3.70) 26 (6.8) 2.17 (1.32-3.59) 1.53 (0.90-2.61) 0.820 Family history of cardiovascular disease No 159 (4.7) 1 1 86 (3.6) 1 1 Yes 36 (6.9) 1.48 (1.04-2.10) 1.33 (0.92-1.91) 39 (6.4) 1.79 (1.24-2.59) 1.48 (1.00-2.17) 0.589 Newborn’s sex Male 118 (5.8) 1 1 65 (4.3) 1 1 Female 77 (4.1) 0.70 (0.53-0.92) 0.68 (0.51-0.91) 60 (4.0) 0.93 (0.66-1.31) 0.92 (0.55-1.49) 0.185 Smoking status Never smoker 127 (5.2) 1 1 88 (4.7) 1 1 Current smoker 41 (3.9) 0.74 (0.52-1.04) 0.77 (0.53-1.10) 26 (3.3) 0.70 (0.46-1.08) 0.82 (0.53-1.29) Ex-smoker 27 (6.3) 1.20 (0.80-1.80) 1.21 (0.80-1.84) 11 (8.8) 0.64 (0.35-1.19) 0.71 (0.37-1.33) 0.327 Prepregnancy BMI (Kg/m 2 ) <25.0 97 (3.3) 1 1 46 (2.4) 1 1 25.0-29.9 54 (7.5) 2.29 (1.65-3.16) 1.94 (1.37-2.76) 40 (5.3) 2.21 (1.46-3.34) 1.78 (1.14-2.77) ≥30 44 (16.4) 4.96 (3.55-6.93) 4.16 (2.87-6.03) 39 (11.4) 4.78 (3.17-7.21) 3.61 (2.29-5.69) 0.704 Weight gain during pregnancy ¥ As recommended 53 (3.7) 1 1 38 (3.4) 1 1 Bellow recommended 36 (3.9) 1.06 (0.70-1.60) 1.07 (0.70-1.63) 27 (3.1) 0.90 (0.55-1.46) 0.91 (0.55-1.49) Above recommended 106 (6.7) 1.79 (1.30-2.48) 1.46 (1.04-2.05) 60 (5.8) 1.70 (1.14-2.53) 1.41 (0.93-2.14) 0.841 108 Most studies that reported an increased risk of later hypertension and cardiovascular outcomes after preeclampsia were retrospective and focus on the occurrence of the outcomes a long time after pregnancy, 4 limiting the ability to study the time frame for development of such outcomes. An association with chronic hypertension shortly after delivery would argue in favour of the continuous nature of the pathological process, once long term effects are know. The aim of this study was to assess the extent to which gestational hypertensive disorders lead to higher blood pressure in women and their offspring as early as 4 years after birth, and to assess the effect of fetal sex on these consequences in a prospectively followed birth cohort. 109 METHODS Study design This study is based on the birth cohort Geração XXI, assembled between 2005 and 2006 at 5 public maternity units covering the metropolitan area of Porto, Portugal. All the maternities corresponded to units with differentiated perinatal support, and were included in a general hospital, except one. Of the invited mothers 91.4% accepted to participate. A total of 8495 mothers, who gave birth to 8647 infants, were enrolled in the cohort. At 4 years of the child’s age, the cohort was re-evaluated. Mothers were invited to participate in a parallel study on women’s cardiovascular health, and 67.3% attended a face-to-face interview and physical examination at the study site. An additional 16.7% provided data by telephone interview, but were not included in the current analysis due to the lack of blood pressure measurement. We excluded 313 participants who had been recruited and evaluated during the first trimester of pregnancy to address specific objectives 18 because the diagnosis and management could be expected to be different in these cases who had prenatal care at the hospital and 3461 multiparous women. Among the primiparae, we excluded 91 with multiple gestations, 85 with chronic hypertension before pregnancy and 39 with missing data on gestational hypertensive disorders. From the remaining, 3070 participated in the follow-up evaluation at the study site. After exclusion of 139 women who were pregnant at the followup visit and 54 with missing data on key variables, 2877 women were available for the current analysis. From those, 99 had a diagnosis of gestational hypertensive disorders during the index pregnancy. Blood pressure measurement in children only started in September 2009. Thus, data on blood pressure among children was available for 2460 consecutively recruited cohort members. Among these, 86 were born from mothers who developed gestational hypertensive disorders (Figure 1). 110 Among the 4506 eligible mothers, those who attended the follow-up re-evaluation were older [mean (standard deviation (SD)): 28.4 (5.4) vs. 26.4 (5.0) years, p<0.001], had a higher socioeconomic position (schooling >12 years: 31.6% vs. 20.3%, p<0.001 and income >1500€/month: 31.6% vs. 20.3%, p<0.001), and were more likely to live with a partner (92.5% vs. 88.6%, p<0.001) than those who did not participate in the re-evaluation, but there were no significant differences between the two groups regarding the prevalence of gestational hypertensive disorders (3.4% vs. 3.3%, p=0.811). The proportion of low birth weight (8.4% vs. 7.2%, p=0.450) and preterm newborns (11.0% vs. 10.2%, p=0.457) was not significantly different among children who participated and those who missed the follow-up reevaluation, and there were no differences regarding the sex of the child (male sex: 51.3% vs. 52.2%, p=0.566). Baseline evaluation At baseline, data on demographic and socioeconomic characteristics, lifestyles, reproductive history and anthropometrics were collected within 72 hours after delivery, during the hospital stay, in a face-to-face interview conducted by trained interviewers using structured questionnaires. Clinical records were also reviewed at birth to retrieve data on complications of pregnancy, birth weight and gestational age of the newborn. Personal history of hypertension was considered present when participants recalled a medical diagnosis of this condition and/or reported antihypertensive drug therapy prescribed specifically for hypertension, before the current pregnancy. Prepregnancy weight was recalled and recorded to the nearest 0.1 Kg and height was measured by the interviewers to the nearest 0.1 cm. When measurement was not possible, height was self-reported as registered in the identity card. The participants’ body mass index (BMI) was categorized according to the standard World Health Organization definition. 19 Weight gain during pregnancy was calculated as the difference between the mother’s 111 reported weight before delivery and prepregnancy weight, and was categorized according to the Institute of Medicine recommendation. 20 Gestational hypertensive disorders were defined by the presence of gestational hypertension or preeclampsia/eclampsia, always considered only when explicitly recorded on obstetrical records as a diagnosis during the index pregnancy. Gestational hypertension was only considered in the absence of a previous diagnosis of chronic hypertension. Newborn’s birth weight was registered to the nearest 1g and gestational age to the nearest 0.1 weeks. Classes of the sex specific adequate birth weight for gestational age were defined according to the Canadian fetal growth standard. 21 Missing data on the questionnaires at baseline were recovered through the review of obstetrical records. The agreement between data collected by questionnaire and abstracted from medical records was good for personal history of hypertension, weight and height, and there was very low inter-rater variability between two independent abstractors. 22 Follow-up evaluation At the cohort’s follow-up, after a median [interquartile range (IQR)] follow-up time of 50 (48-54) months, data were collected by trained interviewers using structured questionnaires, regarding the health of the mother and her child. For the sake of simplicity, 4 years of age will be assumed hereafter. Socioeconomic characteristics, personal and family history of disease, lifestyles and obstetric history of the mother were self-reported. Family history of hypertension was considered present when the women reported to have at least one parent or sibling affected by this condition. Current smokers included both daily and occasional smokers. Ex-smokers did not smoke for at least 6 months. Anthropometrics of mothers and children were measured with the participants wearing light clothing and no footwear. Weight was measured using a digital scale to the nearest 0.1 kg and height measured to the nearest 0.1 cm. The mothers’ BMI was 112 categorized according to the standard World Health Organization definition. 19 Regarding the children, height and BMI was classified according to the age specific percentiles estimated by the United States Center for Disease Control and Prevention. 23 Blood pressure was measured on a single occasion by non-physician trained interviewers. For the mother, two measurements of blood pressure separated by at least 5 minutes were taken with an automatic upper arm blood pressure monitor (OMRON M6 comfort (HEM-7000-E)) after 10-minute rest, on the dominant upper arm resting at the heart level. The mean was calculated and when the difference was larger than 5 mmHg for systolic or diastolic blood pressure a third measurement was taken and the mean of the 2 closest values was considered. Arterial hypertension was defined as systolic and/or diastolic blood pressure ≥140/90 mmHg and/or self-reported antihypertensive drug therapy prescribed specifically for hypertension. Blood pressure of the child was measured with an aneroid sphygmomanometer (Erka Vario DeskModel), with a proper size cuff. Korotkoff phases I and V were used to indicate systolic and diastolic blood pressure. Blood pressure was measured twice, in a seated position with the antecubital fossa supported at heart level, after 5 minutes of rest. The two measurements of blood pressure were separated by at least 5 minutes and when the difference was larger than 5 mmHg for systolic or diastolic blood pressure a third measurement was taken and the mean of the 2 closest values was considered. Systolic and diastolic blood pressure were classified according to the criteria of the American Academy of Pediatrics24 and hypertension was considered as systolic and/or diastolic blood pressure ≥95th percentile for sex, age and height. 21 Statistical analysis Statistical analysis was performed using the software Stata 9.0 (College Station, TX, 2005). To estimate adjusted mean differences in systolic and diastolic blood pressure, for mothers and offspring, according to gestational hypertensive disorders, multiple linear regression was used. Women under anti-hypertensive drug therapy were excluded from this 113 analysis (n=49). No child was under anti-hypertensive drug therapy. The association between gestational hypertensive disorders and the mother’s incidence of hypertension 4 years after giving birth was estimated by crude and adjusted incidence rate ratios (IRR) and respective 95% confidence intervals (95% CI), using Poisson regression. Both for children and mothers, separate analyses were performed for girls and boys, since an interaction term between fetal sex and gestational hypertensive disorders was statistically significant at the 5% level. Ethics The study protocol was approved by the Ethics Committee of Hospital de São João and by the Portuguese Authority of Data Protection. Written informed consent was obtained from all the mothers and children, who gave it through their legal representative. 114 RESULTS Characteristics of the study sample There were no significant differences in socio-demographic characteristics, the number of subsequent births and the smoking status 4 years after delivery between participants with and without gestational hypertensive disorders (Table 1). Mothers who developed those conditions during pregnancy were more likely to report a family history of hypertension, more often gained more weight during pregnancy than recommended and remained more likely to be overweight and obese 4 years after giving birth. The proportion of women who delivered a small for gestational age newborn was higher among those who developed gestational hypertension or preeclampsia/eclampsia during the index pregnancy. Outcomes of the mother In this sample of Portuguese primiparae, approximately 4 years after delivery, the mean (SD) systolic and diastolic blood pressure were significantly higher in mothers with gestational hypertensive disorders who delivered a girl (120.7 (15.7) vs. 106.4 (11.1) mmHg; p<0.001, and 84.7 (12.5) vs. 73.1 (10.8) mmHg; p<0.001, respectively), and among those who delivered a boy (113.5 (12.5) vs. 105.6 (10.8) mmHg; p<0.001, and 79.5 (9.6) vs. 72.9 (9.0) mmHg; p<0.001, respectively). The 4-year crude incidence rate of hypertension per 1000 person-years was much higher in women with gestational hypertension or preeclampsia/eclampsia during pregnancy who gave birth to a female (114.6 vs. 12.2, p<0.001) or to a male newborn (52.7 vs. 11.6, p<0.001) than normotensive women (Figure 2). After adjustment for age, family history of hypertension, BMI at follow-up and smoking status, gestational hypertensive disorders were significantly associated with an increase of systolic blood pressure of 11 and 6 mmHg for women who delivered a female and a male newborn, respectively (p for interaction=0.018). Similarly, gestational hypertension or preeclampsia/eclampsia were significantly associated with an approximately 8 and 5 mmHg 115 higher diastolic blood pressure in mothers of girls and in mothers of boys, respectively (p for interaction=0.033). Four years after delivery, gestational hypertensive disorders were significantly associated with an almost 6-fold increase in the incidence of hypertension among mothers who delivered a girl, and with a 3-fold increase in those who delivered a boy (p for interaction=0.265) (Table 2). Outcomes of the offspring An average of 4 years after birth, the incidence rate of hypertension per 1000 personyears was higher for both girls (50.0; vs. 32.0, p=0.218) and boys (39.1 vs. 28.0, p=0.334) born of women with gestational hypertension or preeclampsia/eclampsia, when compared with those born of normotensive women (Figure 2). After adjustment for classes of birth weight for gestational age, current age and height, gestational hypertensive disorders were associated with a non-significant increase of approximately 2 mmHg on systolic and 0.5 mmHg on diastolic blood pressure, at age 4 among girls. Among boys, gestational hypertensive disorders remained significantly associated with a 2 and 3 mmHg higher systolic and diastolic blood pressure after adjustment (Table 3). 116 DISCUSSION In this sample of Portuguese primiparous women, the development of gestational hypertensive disorders was significantly associated with higher systolic and diastolic blood pressure, an average of 4 years after giving birth. Overall, the 4-year risk of hypertension in women who developed those conditions during pregnancy was almost 6 times higher among mothers who delivered a girl, and 3 times higher among those who delivered a boy. Systolic and diastolic blood pressure at 4 years were higher in boys born of mothers with hypertensive disorders of pregnancy, while no effect was detected among girls. The increase of more than 5 mmHg in both systolic and diastolic blood pressure, among mothers with gestational hypertensive disorders during the index pregnancy, represents a substantial shift to the right in blood pressure distribution. Similar estimates were found in a case-control study, with a mean increase of 7.3mmHg and 4.7mmHg in systolic and diastolic blood pressure, after a average period of 7.8 years from the index pregnancy. 25 In the long term, at least preeclampsia is associated with a 4-fold increased risk of a later diagnosis of hypertension, with over 50% of women with preeclampsia developing chronic hypertension after 14 years.4 Recently, a prospective study in which women were interviewed by telephone 6 to 13 months after delivery reported a strong and positive association between preeclampsia and self-reported chronic hypertension. 26 Our study adds the prospective follow-up with objective measurement of blood pressure of mothers of a population-based birth cohort, showing that as soon as 4 years after delivery, the risk of chronic hypertension was already significantly higher in women with gestational hypertensive disorders. Taken together and considering the overlap of determinants of chronic and gestational hypertensive disorders, these findings question the transient nature of this hypertension. Evidence of sub-clinical maternal endothelial dysfunction after a pregnancy affected by preeclampsia but before the onset of clinical cardiovascular disease supports that a continuous uninterrupted pathological process develops. 13, 27 Heterogenous causes of 117 gestational hypertensive disorders have been proposed, comprising both a background increased vascular risk possibly with a yet unidentified higher blood pressure beforehand and hypertension induced by disturbed perfusion of the placenta. 12 In light of this causal model, the role of an immune response to paternal antigens in a de novo and transient hypertensive state could be expected to occur more often with male babies, resulting in a smaller effect of hypertension in pregnancy on later blood pressure of the mother when the fetus is male, while hypertension in pregnancy with a female fetus might more often only represent the revelation of a previous higher vascular risk. Despite controversial, 28 aggregation of gestational hypertension and preeclampsia/eclampsia into a composite endpoint is an advantage, assuming the two entities could represent different stages of one pathological process and reducing the possibility of misclassification between more specific outcomes. Moreover, studies that analyzed gestational hypertension and preeclampsia described similar associations with chronic hypertension later in life. 29 It has been argued that women with uncomplicated pregnancies have a lower rate of later cardiovascular disease than the general population and are therefore not an appropriate reference category. 12 Thus, we compared the 4-year incidence of hypertension of our unexposed study participants with that from a general population to assess a putative differential selection in exposed and unexposed. In a representative sample of the adult population of Porto, assembled from 1999 to 2003, 30 the 4-year crude incidence rate of hypertension in women aged below 40 years (median age similar to the one of our cohort) was 22.3 per 1000 person-years (data not shown). This rate is considerably higher than that of women with uncomplicated pregnancies in our current sample, confirming the expected healthier average profile of women who achieved successful uncomplicated pregnancies, but it is still much lower than the one estimated for those with gestational hypertensive disorders. We restricted our analysis to primiparae in order to remove the possible effect of a gestational hypertensive disorder occurring in a previous pregnancy. Additionally, achieving 124 21. Kramer MS, Platt RW, Wen SW, et al. A new and improved population-based Canadian reference for birth weight for gestational age. Pediatrics. 2001; 108(2):E35. 22. Alves E, Lunet N, Correia S, Morais V, Azevedo A, Barros H. Inter-rater variability in medical record review and agreement with self-reported data collected by questionnaire. Gaceta Sanitaria. 2011; 25(3):211-219. 23. Kuczmarski RJ OC, Guo SS, et al. 2000 CDC growth charts for the United States: Methods and development. National Center for Health Statistics. Vital Health Stat. 2002; 11 (246). 24. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics. 2004; 114 (2 Suppl 4th Report):555-576. 25. Forest JC, Girouard J, Masse J, et al. Early occurrence of metabolic syndrome after hypertension in pregnancy. Obstet Gynecol. 2005; 105(6):1373-1380. 26. Edlow AG, Srinivas SK, Elovitz MA. Investigating the risk of hypertension shortly after pregnancies complicated by preeclampsia. Am J Obstet Gynecol. 2009; 200(5):e60-62. 27. Chambers JC, Fusi L, Malik IS, Haskard DO, De Swiet M, Kooner JS. Association of maternal endothelial dysfunction with preeclampsia. JAMA. 2001; 285(12):1607-1612. 28. Villar J, Carroli G, Wojdyla D, et al. Preeclampsia, gestational hypertension and intrauterine growth restriction, related or independent conditions? Am J Obstet Gynecol. 2006; 194(4):921-931. 29. Wilson BJ, Watson MS, Prescott GJ, et al. Hypertensive diseases of pregnancy and risk of hypertension and stroke in later life: results from cohort study. BMJ. 2003; 326(7394):845. 30. Santos AC, Severo M, Barros H. Incidence and risk factors for the metabolic syndrome in an urban South European population. Prev Med. 2010; 50(3):99-105. 31. Kotchen JM, Kotchen TA, Cottrill CM, Guthrie GP, Jr., Somes G. Blood pressures of young mothers and their first children 3-6 years following hypertension during pregnancy. J Chronic Dis. 1979; 32(9-10):653-659. 125 32. Palti H, Rothschild E. Blood pressure and growth at 6 years of age among offsprings of mothers with hypertension of pregnancy. Early Hum Dev. 1989; 19(4):263-269. 33. Di Renzo GC, Rosati A, Sarti RD, Cruciani L, Cutuli AM. Does fetal sex affect pregnancy outcome? Gend Med. 2007; 4(1):19-30. 34. Al-Khan A, Aye IL, Barsoum I, et al. IFPA Meeting 2010 Workshops Report II: Placental pathology; trophoblast invasion; fetal sex; parasites and the placenta; decidua and embryonic or fetal loss; trophoblast differentiation and syncytialisation. Placenta. 2011; 32 Suppl 2:S90-99. 35. Macedo ME, Lima MJ, Silva AO, Alcantara P, Ramalhinho V, Carmona J. Prevalence, awareness, treatment and control of hypertension in Portugal: the PAP study. J Hypertens. 2005; 23(9):1661-1666. 36. Chiolero A, Cachat F, Burnier M, Paccaud F, Bovet P. Prevalence of hypertension in schoolchildren based on repeated measurements and association with overweight. J Hypertens. 2007; 25(11):2209-2217. 37. Sorof JM, Lai D, Turner J, Poffenbarger T, Portman RJ. Overweight, ethnicity, and the prevalence of hypertension in school-aged children. Pediatrics. 2004; 113(3 Pt 1):475-482. 126 Table1. Characteristics of primiparae with single pregnancies and no previous chronic hypertension, and their children, according to sex of the offspring and the occurrence of gestational hypertensive disorders. Girls Boys GHD a Normotensive p GHD a Normotensive p Mothers n=39 n=1359 n=60 n=1419 Age at baseline (years), mean (SD) 29.2 (5.9) 28.4 (5.1) 0.391 28.9 (4.9) 28.4 (5.0) 0.405 Education (years), median (IQR) 10 (7-12) 12 (9-16) 0.082 12 (9-16) 12 (9-16) 0.061 Family history of hypertension, n (%) 27 (69.2) 674 (49.6) 0.016 38 (63.3) 687 (48.4) 0.024 Smoking at 4 years, n (%) 0.231 0.195 Never smoker 24 (61.5) 819 (60.3) 42 (70.0) 835 (59.0) Current smoker 6 (15.4) 335 (24.6) 10 (16.7) 371 (26.2) Ex-smoker 9 (23.1) 205 (15.1) 8 (13.3) 211 (14.9) Weight gain during pregnancy b , n (%) 0.151 0.152 Below recommended 7 (21.2) 327 (26.9) 10 (19.6) 338 (26.5) As recommended 9 (27.3) 461 (37.9) 17 (33.3) 503 (39.5) Above recommended 17 (51.5) 427 (35.1) 24 (47.1) 434 (34.0) BMI at 4 years (kg/m 2 ), n (%) <0.001 0.002 <24.9 5 (12.8) 721 (53.1) 19 (31.7) 747 (52.6) 25.0-29.9 15 (38.5) 393 (28.9) 22 (36.7) 428 (30.2) ≥30.0 19 (48.7) 245 (18.0) 19 (31.7) 244 (17.2) At least one birth after the index birth, n (%) 9 (23.1) 272 (20.0) 0.638 9 (15.0) 309 (21.8) 0.210 Children n=35 n=1151 n=51 n=1223 Birth weight for gestational age c , n (%) 0.010 0.369 Small (<10th percentile) 10 (28.6) 180 (15.6) 12 (23.5) 197 (16.1) Adequate (10th - 90 percentile) 22 (62.9) 940 (81.7) 38 (74.5) 1004 (82.1) Large (≥90th percentile) 3 (8.6) 31 (2.7) 1 (2.0) 22 (1.8) Age at follow-up (months), median (IQR) 51 (50-55) 53 (50-59) 0.190 53 (50-58) 51 (50-55) 0.377 BMI, body mass index; GHD, gestational hypertensive disorders; IQR, interquartile range; SD, standard deviation a Gestational hypertension and/or preeclampsia/eclampsia b According to the Institute of Medicine classification (2009) c According to Canadian fetal growth standard 127 Table 2. Mean difference (β) of systolic and diastolic blood pressure (mmHg), and incidence rate ratio (IRR) of hypertension, an average of 4 years after delivery, for mothers who delivered girls (n=1398) and boys (n=1479), according to gestational hypertensive disorders. Women under antihypertensive drug therapy were excluded from the analysis of mean blood pressure. Systolic blood pressure Diastolic blood pressure Hypertension β (95% CI) β (95% CI) IRR (95% CI) Crude Adjusted a Crude Adjusted a Crude Adjusted a Mothers who delivered girls Gestational hypertensive disorders (yes vs. no) 14.89 (10.93 to 18.85) 12.34 (8.57 to 16.11) 11.55 (8.61 to 14.49) 8.52 (5.73 to 11.30) 12.07 (8.82 to 15.32) 9.89 (6.78 to 12.89) Mothers who delivered boys Gestational hypertensive disorders (yes vs. no) 7.71 (4.85 to 10.57) 6.16 (3.37 to 8.94) 6.54 (4.22 to 8.86) 4.80 (2.58 to 7.02) 6.20 (3.82 to 8.58) 4.62 (2.35 to 6.89) 95%CI, 95% confidence interval; β, linear regression coefficient; IRR, incidence rate ratio a Adjusted for mothers’ age at birth, family history of hypertension, body mass index (BMI) at follow-up and smoking status at follow-up 128 Table 3. Mean difference (β) of systolic and diastolic blood pressure (mmHg) for girls (n=1186) and boys (n=1274), at 4 years, according to gestational hypertensive disorders. Systolic blood pressure Diastolic blood pressure β (95% CI) β (95% CI) Crude Adjusted a Crude Adjusted a Girls Gestational hypertensive disorders (yes vs. no) 2.36 (-0.45 to 5.17) 1.93 (-0.71 to 4.56) 0.56 (-2.19 to 3.32) 0.48 (-2.20 to 3.17) Boys Gestational hypertensive disorders (yes vs. no) 2.60 (0.28 to 4.92) 2.29 (0.10 to 4.48) 3.10 (0.83 to 5.38) 3.11 (0.87 to 5.34) 95%CI, 95% confidence interval; β, linear regression coefficient. a Adjusted for childs’ percentile of birth weight for gestational age classes (according to the Canadian fetal growth standard), age (continuous) and height percentile classes at follow-up (according to the CDC criteria) 129 Figure 1. Flowchart for study sample definition. 8495 mothers 4506 mothers and children at baseline 8647 children 313 (3.7%) recruited and evaluated during pregnancy 3461 (40.7%) multiparae 91 (1.1%) multiple gestations 85 (1.0%) with chronic hypertension before pregnancy 39 (0.5%) with missing data on gestational hypertensive disorders 3070 face-to-face follow-up interview 759 telephone follow-up interview 1398 gave birth to a girl 1479 gave birth to a boy 39 with GHD 1359 without GHD 60 with GHD 1419 without GHD 2460 face-to-face follow-up interview (with BP measurement) 713 face-to-face follow-up interview (without BP measurement) 1186 girls 1274 boys 35 with GHD 1151 without GHD 51 with GHD 1223 without GHD Mothers Children 751 telephone follow-up interview 4721 primiparae 130 Figure 2. Incidence rates of hypertension among mothers and children, according to the occurrence of gestational hypertensive disorders (GHD) in the index pregnancy and child’s sex. The symbols represent the incidence rate point estimates and vertical lines the 95% confidence intervals. Hypertension was defined as blood pressure ≥140/90 mmHg and/or or reported antihypertensive drug therapy prescribed specifically for hypertension in women and as systolic and/or diastolic blood pressure ≥95th percentile for sex, age and height in children. 131 PAPER V Alves E, Azevedo A, Correia S, Barros H. Pregnancy and afterwards: opportunities to modify the course of the smoking epidemic in a female population at high risk. [submitted] 132 133 Pregnancy and afterwards: opportunities to modify the course of the smoking epidemic in a female population at high risk Elisabete Alves 1,2 , Ana Azevedo 1,2 , Sofia Correia 1,2 , Henrique Barros 1,2 1 Department of Clinical Epidemiology, Predictive Medicine and Public Health, University of Porto Medical School, Porto, Portugal 2 Institute of Public Health – University of Porto (ISPUP), Porto, Portugal;