TY - JOUR
AU - Bettiol,, Heloisa
AB - Abstract The rate of cesarean delivery (CD) is high in many parts of the world. Birth via CD has been associated with adverse later health outcomes, such as obesity, asthma, and type 1 diabetes mellitus. Few studies have focused on hypertension. We investigated the associations of CD with hypertension, systolic blood pressure (BP), and diastolic BP and tested whether body mass index (BMI; weight (kg)/height (m)2) was a mediator of these associations in a birth cohort (n = 2,020) assembled in 1978–1979 and followed up in 2002–2004 in Ribeirão Preto, Brazil. The CD rate was 32.0%. Hypertension was present in 11.7% of persons born via CD and 7.7% of those born vaginally. Being born by CD increased the odds of hypertension by 51% (odds ratio = 1.51, 95% confidence interval (CI): 1.10, 2.07). After adjustment for confounders, this estimate changed little (odds ratio = 1.49, 95% CI: 1.07, 2.06). In a mediation analysis, odds ratios for the indirect and direct effects were 1.18 (95% CI: 1.11, 1.25) and 1.31 (95% CI: 0.97, 1.65), respectively. CD also had indirect effects on both systolic and diastolic BP via BMI. Our findings suggest that CD is associated with young-adult hypertension and that this association is at least partially mediated by BMI. This has implications for countries struggling with the burden of noncommunicable diseases and where CD rates are high. cesarean delivery, hypertension, life cycle The rate of cesarean delivery (CD) is high in both developed and developing countries. In Brazil, nearly 55% of babies are born via CD (1). According to the World Health Organization, the optimal CD rate is approximately 10%–15% (2). Birth by CD has been associated with later-life health outcomes. In a recent meta-analysis of 28 studies, Kuhle et al. (3) found that children born via CD had a 34% greater risk (95% confidence interval (CI): 1.18, 1.51) of being obese during childhood and adolescence. In another meta-analysis, which included broader age groups, Li et al. (4) found that the association was present within different age strata. Similar findings were reported when highly different socioeconomic settings were compared (5). Other studies, however, did not find an association between birth via CD and obesity (6, 7). Studies have also demonstrated links between birth via CD and chronic diseases such as asthma (8) and type 1 diabetes mellitus (9). To our knowledge, only 1 study to date has examined the association between birth via CD and blood pressure (BP). In that study, systolic BP was 1.15 mm Hg higher among persons born via CD (10) as compared with those born vaginally. After controlling for body mass index (BMI), the association with BP was attenuated, but no formal mediation analysis was performed. The objective of the present work was to investigate the association of CD with BP and hypertension in adulthood, in a sample from a population with high CD rates that varies markedly in terms of indications for elective CD. We also applied mediation analysis to verify whether BMI mediated these associations. METHODS Study design and sample Our data came from a prospective birth cohort study performed in the city of Ribeirão Preto, Brazil. The cohort was established in 1978–1979, when 94.5% of all live births to women resident in the city, among which 6,827 were singleton births, were entered into the study (n = 6,973). In 2002–2004, when the offspring were aged 23–25 years, 6,484 participants were known to be alive, and attempts were made to locate them for a follow-up interview; 5,665 were successfully located. From the list of the identified 5,665 cohort members, a systematic 1-in-3 sample was identified (the first of every 3 names was selected from a list sorted by birth date in each geographical region; if the person was unavailable, the next name down was selected). Replacement occurred in 705 cases because of refusal to participate, incarceration, death after 20 years of age, or failure to attend the interview; 2,063 cohort members were interviewed (11). Figure 1 shows selection of the study sample. The sampling strategy has been described in detail elsewhere (12). The study had approximately 80% power to detect a prevalence ratio of 1.7, assuming that the event had a 7% prevalence in the control group, with a 5% probability of type I error. Figure 1. View largeDownload slide Selection of persons from a population-based cohort born in 1978–1979 in Ribeirão Preto, Brazil, for an analysis of cesarean delivery and early adult hypertension. Figure 1. View largeDownload slide Selection of persons from a population-based cohort born in 1978–1979 in Ribeirão Preto, Brazil, for an analysis of cesarean delivery and early adult hypertension. Instruments and variables Information on maternal characteristics was collected by interview soon after delivery. We extracted data on the following variables: occupation of the family head (nonskilled manual, semiskilled manual, skilled manual, or nonmanual labor), based on the International Classification of Occupations (13); maternal age at delivery (≤19, 20–34, or ≥35 years); maternal duration of schooling (≤4, 5–8, or ≥9 years); number of cigarettes smoked per day during pregnancy (0, 1–10, or >10); hypertension (defined below) during pregnancy (yes/no); type of delivery (vaginal/CD); and gestational age (completed weeks), based on the date of the last menstrual period. Gestational age at delivery (weeks) was used as a continuous variable. Trained personnel measured the newborn’s weight and length immediately after delivery. The babies were weighed naked in weekly-calibrated scales with 10-g precision. Two staff members measured the newborn’s length using a neonatometer, with the baby laid in a supine position (14). As adults, participants responded to a general questionnaire. Sixty-eight percent were still living with their birth family. The current occupation of the head of the participant’s family was classified into one of 4 categories (nonmanual, skilled manual, semiskilled manual, or unskilled manual). We administered the short version of the International Physical Activity Questionnaire (15); participants were classified as sedentary, sufficiently active, or active based on the instrument guidelines. A food frequency questionnaire, validated for the Brazilian population (16) and applied by a nutritionist, ascertained information about habitual intake of 75 food items over the past 12 months. The nutritional value of the participants’ diets was analyzed with the software DietSys 4.0 (National Cancer Institute, Bethesda, Maryland); Brazilian foods and food preparations complementary to the DietSys program were added when complete data about them were available from Brazilian tables (17). Participants’ daily alcohol consumption was considered as none, low (≤31 g), or high (>31 g). Sodium intake was considered as only sodium present in foods consumed and was categorized as <2,000 mg/day, 2,000–2,500 mg/day, or >2,500 mg/day. The number of cigarettes smoked per day was categorized as 0, 1–10, or >10. Ethnicity was categorized as black or nonblack on the basis of self-reported skin color. BP was measured in triplicate (18) using a 740 Omron digital sphygmomanometer (OMRON Healthcare Europe B.V., Hoofdorp, the Netherlands) with a cuff size appropriate for the participant’s arm circumference. One person took all BP measurements at 15-minute intervals, with the participant resting in a sitting position with the left arm at the height of the heart. The mean of the last 2 measurements was used in the analysis. Persons with a systolic BP at or above 140 mm Hg and/or a diastolic BP at or above 90 mm Hg were considered hypertensive, as were those who were taking medication for hypertension. Weight was measured with an adult electronic scale (Filizola, São Paulo, Brazil) with 100-g precision. Height was measured to the nearest centimeter using a stadiometer made of wood and laminate that was resistant to deformation, according to standard techniques (19). The medical directors of all maternity hospitals approved the original cohort study, and the mothers gave their verbal consent to participate in the study. The adult wave of follow-up was approved by the Research Ethics Committee of the university hospital of the Faculty of Medicine of Ribeirão Preto, University of São Paulo, and all subjects gave written informed consent to participate. Data processing and analysis Assuming that data were missing at random, we dealt with missing values in early-life variables using multiple imputation by chained equations (20). Estimates were averaged over 20 imputed data sets. Data missing in the follow-up wave were excluded (43 cases), leaving 2,020 observations for the regression and mediation analysis. We computed BMI as weight (kg)/height (m)2. Height and BMI were standardized for sex. Mean values and standard deviations were used to characterize continuous variables. Relative frequencies were used to describe data according to the exposure (vaginal delivery or CD) and the outcome (hypertension in young adulthood). Groups were compared through χ2 tests or t tests. We performed logistic regression to estimate odds ratios for the association between CD and hypertension. We also built linear regression models for the associations of CD with systolic and diastolic BP. In our sample, 18 people were using medication for hypertension. Two participants were using 2 types of medicine. According to Wald et al. (21), each medication is able to reduce systolic BP by 8 mm Hg and diastolic BP by 5 mm Hg. Therefore, we added these amounts to the measured BPs of persons who were using 1 medication and twice these amounts to those who were using 2 types of medication. There was no effect of interaction between CD and sex on hypertension. Since persons who participated in the follow-up were slightly different from those who did not participate, we used inverse probability weighting to deal with possible selection bias. The probability of participation in the follow-up phase was calculated in a logistic regression model using all variables that were associated with loss to follow-up from birth to adulthood as predictors (22) (sex, gestational age at delivery, maternal schooling, occupation of the head of the family at birth, maternal age, and maternal smoking). We then generated a weight variable that was the inverse of this probability. The distribution of this variable suggested no evidence of inflation (normal distribution; mean = 3.24; range, 2.14–5.40). Unstabilized weights were used, since there were no extremely high weights. Covariates were selected with the aid of a directed acyclic graph (23). Variable categories were ordered according to their specific temporal relationships. Identification of temporal order facilitates understanding of links between variables. Type of delivery was linked to anthropometric variables, since it has been associated with growth patterns. The resulting graphic of hypothesized associations was built in a browser-based environment (http://www.dagitty.net) (see Web Figure 1, available at https://academic.oup.com/aje). The backdoor criterion was used to choose the minimum set of variables needed to adjust for confounders (Web Table 1). According to this criterion, the minimum set of adjustment factors for confounding should include all variables that were able to block all open backdoor paths from the exposure to the outcome and should not include any descendants of the exposure (24). In the mediation analysis, adjustment was made for exposure-outcome, exposure-mediator, and mediator-outcome confounders identified with the aid of the directed acyclic graph. There was no interaction between the exposure (CD) and the mediator (BMI). Because there was no interaction, the controlled direct effect and the natural direct effect were the same. There was no evidence of a nonlinear association between BMI and the probability of hypertension or between BMI and systolic or diastolic BP. The natural direct effect shows how much hypertension would change when comparing CD with vaginal delivery while keeping BMI at the level at which it would have been among those born vaginally. The natural indirect effect depicts how much hypertension would change, on average, among persons born by CD while changing BMI from the level at which it would have been if the adult had been born vaginally to the level at which it would have been if the adult had been born by CD. The assumptions used to estimate average natural direct and indirect effects were: 1) there are no unmeasured exposure-outcome or exposure-mediator confounders, apart from those used in the adjusted models; 2) there are no unmeasured mediator-outcome confounders, apart from type of delivery and the confounders included in the adjusted models; and 3) there is no mediator-outcome confounder affected by the type of delivery. To test the mediation effect of BMI as a continuous variable on the association between CD and hypertension or systolic or diastolic BP, we used the STATA (StataCorp LLC, College Station, Texas) command “paramed” (25), which estimates direct and indirect effects on the basis of counterfactual definitions. Since “paramed” does not support weighted analysis, we additionally chose to calculate total and direct effects through weighted conventional regression. The STATA 14.0 statistical package was used for all analyses. The statistical significance level was set at 0.05. RESULTS The studied sample differed from the original birth cohort in some respects. Participants who were male, were born preterm, were born to mothers with a low level of schooling, were smokers, or were younger or whose family heads were engaged in manual occupations were less likely to be followed up (Web Table 2). The rate of birth via CD was 32.0% (n = 661). At follow-up, 8.9% (n = 184) of the studied participants had hypertension. The mean systolic BP was 126.8 (standard deviation (SD), 13.1) mm Hg among men and 109.6 (SD, 11.3) mm Hg among women. The mean diastolic BP was 74.0 (SD, 9.0) mm Hg among men and 68.0 (SD, 8.3) mm Hg among women. The mean BMI and height were 25 (SD, 4.4) and 176.0 (SD, 6.5) cm, respectively, among men and 23.6 (SD, 5.1) and 162.6 (SD, 6.4) cm, respectively, among women. Table 1 shows that 11.7% of participants born via CD were hypertensive; the proportion with hypertension among those delivered vaginally was 7.7%. CD was associated with higher parental occupational category at the time of the birth of subjects and in adulthood; with higher maternal schooling and age; with more frequent nonblack skin color; with higher offspring BMI at birth and in adulthood; and with higher adult schooling. Hypertensive participants were more likely to be male, to have been born to younger mothers, to have higher adult BMI and height, and to report greater alcohol consumption and lower physical activity. Table 1. Selected Characteristics (%) of a Population-Based Cohort Born in 1978–1979 in Ribeirão Preto, Brazil, by Type of Delivery and by Hypertension Status in Adulthood Variable No. of Participantsa,b Type of Delivery Hypertension Status Vaginal (n = 1,402) Cesarean (n = 661) Normotensive (n = 1,876) Hypertensive (n = 184) Occupation of the family head at birth  Nonskilled manual labor 339 23.0 13.5 20.0 19.0  Semiskilled manual labor 1,265 62.7 64.0 63.4 60.9  Nonmanual labor + skilled manual labor 399 14.3 22.5 16.6 20.1 Maternal duration of schooling, years  ≤4 920 49.1 37.9 46.3 37.6  5–8 557 27.2 28.2 27.0 32.6  9–11 331 15.1 19.0 16.2 18.0  ≥12 215 8.6 14.9 10.5 11.8 Maternal age at birth, years  ≤19 254 14.7 7.6 12.8 8.7  20–34 1,626 78.4 81.0 79.4 77.6  ≥35 171 6.9 11.4 7.8 13.7 Male offspring sex 995 48.2 48.3 44.7 84.2 Nonblack offspring skin color 1,367 63.5 72.2 66.8 60.9 Maternal smoking during pregnancy, cigarettes/day  0 (none) 1,593 78.1 81.6 79.1 81.0  1–10 312 16.0 14.5 15.6 14.0  >10 106 5.9 3.9 5.3 5.0 Maternal hypertension (yes)c 30 1.3 1.8 1.4 1.6 Child’s BMId at birthe 2,052 13.3 (1.4) 13.8 (1.5) 13.5 (1.5) 13.4 (1.4) Birth length, cme 2,052 49.1 (2.2) 49.2 (2.4) 49.1 (2.2) 49.3 (2.3) Gestational age at delivery, weekse 2,063 39.2 (1.8) 39.3 (1.8) 39.3 (1.8) 39.1 (1.9) Birth via cesarean delivery 2,063 31.9 46.0 Adult subject’s duration of schooling, years  ≤4 58 3.1 2.3 2.5 6.0  5–8 262 14.0 10.0 12.6 13.6  9–11 1,039 52.8 45.2 50.5 48.9  ≥12 704 30.1 42.5 34.4 31.5 Occupation of family head in adulthood  Nonskilled manual labor 358 18.4 15.2 17.4 17.6  Semiskilled manual labor 434 22.5 18.1 21.5 17.6  Skilled manual labor 570 28.2 26.6 27.7 28.0  Nonmanual labor 696 30.8 40.2 33.5 36.8 Subject’s smoking in adulthood, cigarettes/day  0 (none) 1,709 82.0 84.6 83.0 81.0  1–10 214 10.8 9.5 10.6 8.7  >10 140 7.2 5.9 6.4 10.3 Subject’s alcohol intake in adulthood  None 564 28.5 25.5 28.4 17.9  Low (≤31 g/day) 1,069 51.1 54.5 52.5 49.5  High (>31 g/day) 416 20.4 20.0 19.1 32.6 Subject’s sodium intake in adulthood, mg/day  <2,000 331 16.3 15.4 16.2 14.7  2,000–2,500 984 47.7 47.8 48.4 41.9  >2,500 748 36.0 36.8 35.5 43.5 Subject’s physical activity level in adulthoodf  Sedentary (≤600 MET-minutes/week) 1,022 50.6 47.7 50.5 41.3  Sufficiently active (>600–≤1,500 MET-minutes/week) 395 19.3 18.9 19.1 19.6  Active (>1,500 MET-minutes/week) 640 30.1 33.3 30.4 39.1 Hypertension in adulthood (yes) 2,060 7.7 11.7 Systolic BP, mm Hge 2,060 118.0 (14.9) 118.3 (15.4) 115.3 (12.3) 144.6 (13.0) Diastolic BP, mm Hge 2,060 70.8 (9.1) 71.6 (9.6) 69.5 (7.9) 84.9 (9.4) Subject’s BMI in adulthoode 2,056 23.9 (4.6) 25.0 (5.3) 23.9 (4.5) 28.1 (5.8) Variable No. of Participantsa,b Type of Delivery Hypertension Status Vaginal (n = 1,402) Cesarean (n = 661) Normotensive (n = 1,876) Hypertensive (n = 184) Occupation of the family head at birth  Nonskilled manual labor 339 23.0 13.5 20.0 19.0  Semiskilled manual labor 1,265 62.7 64.0 63.4 60.9  Nonmanual labor + skilled manual labor 399 14.3 22.5 16.6 20.1 Maternal duration of schooling, years  ≤4 920 49.1 37.9 46.3 37.6  5–8 557 27.2 28.2 27.0 32.6  9–11 331 15.1 19.0 16.2 18.0  ≥12 215 8.6 14.9 10.5 11.8 Maternal age at birth, years  ≤19 254 14.7 7.6 12.8 8.7  20–34 1,626 78.4 81.0 79.4 77.6  ≥35 171 6.9 11.4 7.8 13.7 Male offspring sex 995 48.2 48.3 44.7 84.2 Nonblack offspring skin color 1,367 63.5 72.2 66.8 60.9 Maternal smoking during pregnancy, cigarettes/day  0 (none) 1,593 78.1 81.6 79.1 81.0  1–10 312 16.0 14.5 15.6 14.0  >10 106 5.9 3.9 5.3 5.0 Maternal hypertension (yes)c 30 1.3 1.8 1.4 1.6 Child’s BMId at birthe 2,052 13.3 (1.4) 13.8 (1.5) 13.5 (1.5) 13.4 (1.4) Birth length, cme 2,052 49.1 (2.2) 49.2 (2.4) 49.1 (2.2) 49.3 (2.3) Gestational age at delivery, weekse 2,063 39.2 (1.8) 39.3 (1.8) 39.3 (1.8) 39.1 (1.9) Birth via cesarean delivery 2,063 31.9 46.0 Adult subject’s duration of schooling, years  ≤4 58 3.1 2.3 2.5 6.0  5–8 262 14.0 10.0 12.6 13.6  9–11 1,039 52.8 45.2 50.5 48.9  ≥12 704 30.1 42.5 34.4 31.5 Occupation of family head in adulthood  Nonskilled manual labor 358 18.4 15.2 17.4 17.6  Semiskilled manual labor 434 22.5 18.1 21.5 17.6  Skilled manual labor 570 28.2 26.6 27.7 28.0  Nonmanual labor 696 30.8 40.2 33.5 36.8 Subject’s smoking in adulthood, cigarettes/day  0 (none) 1,709 82.0 84.6 83.0 81.0  1–10 214 10.8 9.5 10.6 8.7  >10 140 7.2 5.9 6.4 10.3 Subject’s alcohol intake in adulthood  None 564 28.5 25.5 28.4 17.9  Low (≤31 g/day) 1,069 51.1 54.5 52.5 49.5  High (>31 g/day) 416 20.4 20.0 19.1 32.6 Subject’s sodium intake in adulthood, mg/day  <2,000 331 16.3 15.4 16.2 14.7  2,000–2,500 984 47.7 47.8 48.4 41.9  >2,500 748 36.0 36.8 35.5 43.5 Subject’s physical activity level in adulthoodf  Sedentary (≤600 MET-minutes/week) 1,022 50.6 47.7 50.5 41.3  Sufficiently active (>600–≤1,500 MET-minutes/week) 395 19.3 18.9 19.1 19.6  Active (>1,500 MET-minutes/week) 640 30.1 33.3 30.4 39.1 Hypertension in adulthood (yes) 2,060 7.7 11.7 Systolic BP, mm Hge 2,060 118.0 (14.9) 118.3 (15.4) 115.3 (12.3) 144.6 (13.0) Diastolic BP, mm Hge 2,060 70.8 (9.1) 71.6 (9.6) 69.5 (7.9) 84.9 (9.4) Subject’s BMI in adulthoode 2,056 23.9 (4.6) 25.0 (5.3) 23.9 (4.5) 28.1 (5.8) Abbreviations: BMI, body mass index; BP, blood pressure; MET, metabolic equivalent of task. a Unweighted number. b Numbers for some variables do not add up to the total because of missing values. c Hypertension was defined as systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg or use of medication for hypertension. d Weight (kg)/height (m)2. e Values are expressed as mean (standard deviation). f Physical activity was assessed using the short version of the International Physical Activity Questionnaire (15). Table 1. Selected Characteristics (%) of a Population-Based Cohort Born in 1978–1979 in Ribeirão Preto, Brazil, by Type of Delivery and by Hypertension Status in Adulthood Variable No. of Participantsa,b Type of Delivery Hypertension Status Vaginal (n = 1,402) Cesarean (n = 661) Normotensive (n = 1,876) Hypertensive (n = 184) Occupation of the family head at birth  Nonskilled manual labor 339 23.0 13.5 20.0 19.0  Semiskilled manual labor 1,265 62.7 64.0 63.4 60.9  Nonmanual labor + skilled manual labor 399 14.3 22.5 16.6 20.1 Maternal duration of schooling, years  ≤4 920 49.1 37.9 46.3 37.6  5–8 557 27.2 28.2 27.0 32.6  9–11 331 15.1 19.0 16.2 18.0  ≥12 215 8.6 14.9 10.5 11.8 Maternal age at birth, years  ≤19 254 14.7 7.6 12.8 8.7  20–34 1,626 78.4 81.0 79.4 77.6  ≥35 171 6.9 11.4 7.8 13.7 Male offspring sex 995 48.2 48.3 44.7 84.2 Nonblack offspring skin color 1,367 63.5 72.2 66.8 60.9 Maternal smoking during pregnancy, cigarettes/day  0 (none) 1,593 78.1 81.6 79.1 81.0  1–10 312 16.0 14.5 15.6 14.0  >10 106 5.9 3.9 5.3 5.0 Maternal hypertension (yes)c 30 1.3 1.8 1.4 1.6 Child’s BMId at birthe 2,052 13.3 (1.4) 13.8 (1.5) 13.5 (1.5) 13.4 (1.4) Birth length, cme 2,052 49.1 (2.2) 49.2 (2.4) 49.1 (2.2) 49.3 (2.3) Gestational age at delivery, weekse 2,063 39.2 (1.8) 39.3 (1.8) 39.3 (1.8) 39.1 (1.9) Birth via cesarean delivery 2,063 31.9 46.0 Adult subject’s duration of schooling, years  ≤4 58 3.1 2.3 2.5 6.0  5–8 262 14.0 10.0 12.6 13.6  9–11 1,039 52.8 45.2 50.5 48.9  ≥12 704 30.1 42.5 34.4 31.5 Occupation of family head in adulthood  Nonskilled manual labor 358 18.4 15.2 17.4 17.6  Semiskilled manual labor 434 22.5 18.1 21.5 17.6  Skilled manual labor 570 28.2 26.6 27.7 28.0  Nonmanual labor 696 30.8 40.2 33.5 36.8 Subject’s smoking in adulthood, cigarettes/day  0 (none) 1,709 82.0 84.6 83.0 81.0  1–10 214 10.8 9.5 10.6 8.7  >10 140 7.2 5.9 6.4 10.3 Subject’s alcohol intake in adulthood  None 564 28.5 25.5 28.4 17.9  Low (≤31 g/day) 1,069 51.1 54.5 52.5 49.5  High (>31 g/day) 416 20.4 20.0 19.1 32.6 Subject’s sodium intake in adulthood, mg/day  <2,000 331 16.3 15.4 16.2 14.7  2,000–2,500 984 47.7 47.8 48.4 41.9  >2,500 748 36.0 36.8 35.5 43.5 Subject’s physical activity level in adulthoodf  Sedentary (≤600 MET-minutes/week) 1,022 50.6 47.7 50.5 41.3  Sufficiently active (>600–≤1,500 MET-minutes/week) 395 19.3 18.9 19.1 19.6  Active (>1,500 MET-minutes/week) 640 30.1 33.3 30.4 39.1 Hypertension in adulthood (yes) 2,060 7.7 11.7 Systolic BP, mm Hge 2,060 118.0 (14.9) 118.3 (15.4) 115.3 (12.3) 144.6 (13.0) Diastolic BP, mm Hge 2,060 70.8 (9.1) 71.6 (9.6) 69.5 (7.9) 84.9 (9.4) Subject’s BMI in adulthoode 2,056 23.9 (4.6) 25.0 (5.3) 23.9 (4.5) 28.1 (5.8) Variable No. of Participantsa,b Type of Delivery Hypertension Status Vaginal (n = 1,402) Cesarean (n = 661) Normotensive (n = 1,876) Hypertensive (n = 184) Occupation of the family head at birth  Nonskilled manual labor 339 23.0 13.5 20.0 19.0  Semiskilled manual labor 1,265 62.7 64.0 63.4 60.9  Nonmanual labor + skilled manual labor 399 14.3 22.5 16.6 20.1 Maternal duration of schooling, years  ≤4 920 49.1 37.9 46.3 37.6  5–8 557 27.2 28.2 27.0 32.6  9–11 331 15.1 19.0 16.2 18.0  ≥12 215 8.6 14.9 10.5 11.8 Maternal age at birth, years  ≤19 254 14.7 7.6 12.8 8.7  20–34 1,626 78.4 81.0 79.4 77.6  ≥35 171 6.9 11.4 7.8 13.7 Male offspring sex 995 48.2 48.3 44.7 84.2 Nonblack offspring skin color 1,367 63.5 72.2 66.8 60.9 Maternal smoking during pregnancy, cigarettes/day  0 (none) 1,593 78.1 81.6 79.1 81.0  1–10 312 16.0 14.5 15.6 14.0  >10 106 5.9 3.9 5.3 5.0 Maternal hypertension (yes)c 30 1.3 1.8 1.4 1.6 Child’s BMId at birthe 2,052 13.3 (1.4) 13.8 (1.5) 13.5 (1.5) 13.4 (1.4) Birth length, cme 2,052 49.1 (2.2) 49.2 (2.4) 49.1 (2.2) 49.3 (2.3) Gestational age at delivery, weekse 2,063 39.2 (1.8) 39.3 (1.8) 39.3 (1.8) 39.1 (1.9) Birth via cesarean delivery 2,063 31.9 46.0 Adult subject’s duration of schooling, years  ≤4 58 3.1 2.3 2.5 6.0  5–8 262 14.0 10.0 12.6 13.6  9–11 1,039 52.8 45.2 50.5 48.9  ≥12 704 30.1 42.5 34.4 31.5 Occupation of family head in adulthood  Nonskilled manual labor 358 18.4 15.2 17.4 17.6  Semiskilled manual labor 434 22.5 18.1 21.5 17.6  Skilled manual labor 570 28.2 26.6 27.7 28.0  Nonmanual labor 696 30.8 40.2 33.5 36.8 Subject’s smoking in adulthood, cigarettes/day  0 (none) 1,709 82.0 84.6 83.0 81.0  1–10 214 10.8 9.5 10.6 8.7  >10 140 7.2 5.9 6.4 10.3 Subject’s alcohol intake in adulthood  None 564 28.5 25.5 28.4 17.9  Low (≤31 g/day) 1,069 51.1 54.5 52.5 49.5  High (>31 g/day) 416 20.4 20.0 19.1 32.6 Subject’s sodium intake in adulthood, mg/day  <2,000 331 16.3 15.4 16.2 14.7  2,000–2,500 984 47.7 47.8 48.4 41.9  >2,500 748 36.0 36.8 35.5 43.5 Subject’s physical activity level in adulthoodf  Sedentary (≤600 MET-minutes/week) 1,022 50.6 47.7 50.5 41.3  Sufficiently active (>600–≤1,500 MET-minutes/week) 395 19.3 18.9 19.1 19.6  Active (>1,500 MET-minutes/week) 640 30.1 33.3 30.4 39.1 Hypertension in adulthood (yes) 2,060 7.7 11.7 Systolic BP, mm Hge 2,060 118.0 (14.9) 118.3 (15.4) 115.3 (12.3) 144.6 (13.0) Diastolic BP, mm Hge 2,060 70.8 (9.1) 71.6 (9.6) 69.5 (7.9) 84.9 (9.4) Subject’s BMI in adulthoode 2,056 23.9 (4.6) 25.0 (5.3) 23.9 (4.5) 28.1 (5.8) Abbreviations: BMI, body mass index; BP, blood pressure; MET, metabolic equivalent of task. a Unweighted number. b Numbers for some variables do not add up to the total because of missing values. c Hypertension was defined as systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg or use of medication for hypertension. d Weight (kg)/height (m)2. e Values are expressed as mean (standard deviation). f Physical activity was assessed using the short version of the International Physical Activity Questionnaire (15). CD was associated with hypertension in both unadjusted (odds ratio (OR) = 1.51, 95% CI: 1.10, 2.07) and adjusted (OR = 1.49, 95% CI: 1.07, 2.06) analyses. However, CD was not associated with either systolic or diastolic BP, either before or after adjustment for confounding (Table 2). Table 2. Crude and Adjusted Odds Ratios for Hypertension by Type of Delivery and Mediation of the Association Between Cesarean Delivery and Hypertension by Adulthood Body Mass Index in a Population-Based Cohort (n = 2,020) Born in 1978–1979 in Ribeirão Preto, Brazil Modela OR β 95% CI P Value Risk of hypertension  Unadjusted and weightedb 1.51 1.10, 2.07 0.011  Adjustedc and weightedb (total effect) 1.49 1.07, 2.06 0.017  Also adjustedd for the mediator BMI and weightedb (direct effect) 1.27 0.90, 1.81 0.179 Mediation analysis—adjustedd and unweightede  Direct effect 1.31 0.97, 1.65 0.116  Indirect effect 1.18 1.11, 1.25 <0.001  Total effect 1.55 1.20, 1.89 0.013 Linear association with systolic BP  Unadjusted and weightedb −0.16 −1.57, 1.26 0.828  Adjustedc and weightedb (total effect) −0.03 −1.48, 1.42 0.967  Also adjustedd for the mediator BMI and weightedb (direct effect) −1.14 −2.47, 0.20 0.096 Mediation analysis—adjustedd and unweightede  Direct effect −1.19 −2.49, 0.12 0.074  Indirect effect 1.19 0.72, 1.67 <0.001  Total effect 0.01 −1.36, 1.38 0.991 Linear association with diastolic BP  Unadjusted and weightedb 0.47 −0.39, 1.34 0.284  Adjustedc and weightedb (total effect) 0.53 −0.35, 1.40 0.241  Also adjustedd for the mediator BMI and weightedb (direct effect) −0.50 −1.26, 0.26 0.200 Mediation analysis—adjustedd and unweightede  Direct effect −0.38 −1.14, 0.37 0.321  Indirect effect 1.08 0.66, 1.50 <0.001  Total effect 0.70 −0.16, 1.55 0.110 Modela OR β 95% CI P Value Risk of hypertension  Unadjusted and weightedb 1.51 1.10, 2.07 0.011  Adjustedc and weightedb (total effect) 1.49 1.07, 2.06 0.017  Also adjustedd for the mediator BMI and weightedb (direct effect) 1.27 0.90, 1.81 0.179 Mediation analysis—adjustedd and unweightede  Direct effect 1.31 0.97, 1.65 0.116  Indirect effect 1.18 1.11, 1.25 <0.001  Total effect 1.55 1.20, 1.89 0.013 Linear association with systolic BP  Unadjusted and weightedb −0.16 −1.57, 1.26 0.828  Adjustedc and weightedb (total effect) −0.03 −1.48, 1.42 0.967  Also adjustedd for the mediator BMI and weightedb (direct effect) −1.14 −2.47, 0.20 0.096 Mediation analysis—adjustedd and unweightede  Direct effect −1.19 −2.49, 0.12 0.074  Indirect effect 1.19 0.72, 1.67 <0.001  Total effect 0.01 −1.36, 1.38 0.991 Linear association with diastolic BP  Unadjusted and weightedb 0.47 −0.39, 1.34 0.284  Adjustedc and weightedb (total effect) 0.53 −0.35, 1.40 0.241  Also adjustedd for the mediator BMI and weightedb (direct effect) −0.50 −1.26, 0.26 0.200 Mediation analysis—adjustedd and unweightede  Direct effect −0.38 −1.14, 0.37 0.321  Indirect effect 1.08 0.66, 1.50 <0.001  Total effect 0.70 −0.16, 1.55 0.110 Abbreviations: BMI, body mass index; BP, blood pressure; CI, confidence interval; OR, odds ratio. a All estimates were obtained by averaging over 20 multiply imputed data sets. b Estimates were weighted by the inverse probability of taking part in follow-up in early adulthood using sex, gestational age, maternal schooling, occupation of the head of the family at birth, maternal age, and maternal smoking. c Adjusted for confounders of the exposure-outcome association (occupation of the head of the family at birth, maternal schooling, maternal smoking, maternal hypertension, maternal age, birth BMI z score, birth length z score, and gestational age). d Adjusted for confounders of the exposure-outcome association (see above) and for confounders of the exposure-mediator association (birth BMI z score, birth length z score, gestational age, occupation of the head of the family, and maternal smoking) and of the mediator-outcome association (occupation of the head of the family, alcohol consumption, schooling, smoking, and physical activity in adulthood). e Estimates obtained via the “paramed” command in STATA (25) were unweighted because this routine does not currently allow for the use of inverse probability weights. Table 2. Crude and Adjusted Odds Ratios for Hypertension by Type of Delivery and Mediation of the Association Between Cesarean Delivery and Hypertension by Adulthood Body Mass Index in a Population-Based Cohort (n = 2,020) Born in 1978–1979 in Ribeirão Preto, Brazil Modela OR β 95% CI P Value Risk of hypertension  Unadjusted and weightedb 1.51 1.10, 2.07 0.011  Adjustedc and weightedb (total effect) 1.49 1.07, 2.06 0.017  Also adjustedd for the mediator BMI and weightedb (direct effect) 1.27 0.90, 1.81 0.179 Mediation analysis—adjustedd and unweightede  Direct effect 1.31 0.97, 1.65 0.116  Indirect effect 1.18 1.11, 1.25 <0.001  Total effect 1.55 1.20, 1.89 0.013 Linear association with systolic BP  Unadjusted and weightedb −0.16 −1.57, 1.26 0.828  Adjustedc and weightedb (total effect) −0.03 −1.48, 1.42 0.967  Also adjustedd for the mediator BMI and weightedb (direct effect) −1.14 −2.47, 0.20 0.096 Mediation analysis—adjustedd and unweightede  Direct effect −1.19 −2.49, 0.12 0.074  Indirect effect 1.19 0.72, 1.67 <0.001  Total effect 0.01 −1.36, 1.38 0.991 Linear association with diastolic BP  Unadjusted and weightedb 0.47 −0.39, 1.34 0.284  Adjustedc and weightedb (total effect) 0.53 −0.35, 1.40 0.241  Also adjustedd for the mediator BMI and weightedb (direct effect) −0.50 −1.26, 0.26 0.200 Mediation analysis—adjustedd and unweightede  Direct effect −0.38 −1.14, 0.37 0.321  Indirect effect 1.08 0.66, 1.50 <0.001  Total effect 0.70 −0.16, 1.55 0.110 Modela OR β 95% CI P Value Risk of hypertension  Unadjusted and weightedb 1.51 1.10, 2.07 0.011  Adjustedc and weightedb (total effect) 1.49 1.07, 2.06 0.017  Also adjustedd for the mediator BMI and weightedb (direct effect) 1.27 0.90, 1.81 0.179 Mediation analysis—adjustedd and unweightede  Direct effect 1.31 0.97, 1.65 0.116  Indirect effect 1.18 1.11, 1.25 <0.001  Total effect 1.55 1.20, 1.89 0.013 Linear association with systolic BP  Unadjusted and weightedb −0.16 −1.57, 1.26 0.828  Adjustedc and weightedb (total effect) −0.03 −1.48, 1.42 0.967  Also adjustedd for the mediator BMI and weightedb (direct effect) −1.14 −2.47, 0.20 0.096 Mediation analysis—adjustedd and unweightede  Direct effect −1.19 −2.49, 0.12 0.074  Indirect effect 1.19 0.72, 1.67 <0.001  Total effect 0.01 −1.36, 1.38 0.991 Linear association with diastolic BP  Unadjusted and weightedb 0.47 −0.39, 1.34 0.284  Adjustedc and weightedb (total effect) 0.53 −0.35, 1.40 0.241  Also adjustedd for the mediator BMI and weightedb (direct effect) −0.50 −1.26, 0.26 0.200 Mediation analysis—adjustedd and unweightede  Direct effect −0.38 −1.14, 0.37 0.321  Indirect effect 1.08 0.66, 1.50 <0.001  Total effect 0.70 −0.16, 1.55 0.110 Abbreviations: BMI, body mass index; BP, blood pressure; CI, confidence interval; OR, odds ratio. a All estimates were obtained by averaging over 20 multiply imputed data sets. b Estimates were weighted by the inverse probability of taking part in follow-up in early adulthood using sex, gestational age, maternal schooling, occupation of the head of the family at birth, maternal age, and maternal smoking. c Adjusted for confounders of the exposure-outcome association (occupation of the head of the family at birth, maternal schooling, maternal smoking, maternal hypertension, maternal age, birth BMI z score, birth length z score, and gestational age). d Adjusted for confounders of the exposure-outcome association (see above) and for confounders of the exposure-mediator association (birth BMI z score, birth length z score, gestational age, occupation of the head of the family, and maternal smoking) and of the mediator-outcome association (occupation of the head of the family, alcohol consumption, schooling, smoking, and physical activity in adulthood). e Estimates obtained via the “paramed” command in STATA (25) were unweighted because this routine does not currently allow for the use of inverse probability weights. The mediation analysis revealed significant indirect (OR = 1.18, 95% CI: 1.11, 1.25) and total (OR = 1.55, 95% CI: 1.20, 1.89) effects of CD on hypertension, suggesting that BMI is a pathway through which type of delivery is associated with adult hypertension. The direct effect was nonsignificant, but since the point estimate was higher (OR = 1.31, 95% CI: 0.97, 1.65) than that for the indirect effect, it is not possible to discount the possibility that the direct effect could be higher than the indirect effect (Table 2). In the mediation analysis of the association of CD with systolic or diastolic BP, CD had a positive indirect effect on both systolic BP (β = 1.19 mm Hg, 95% CI: 0.72, 1.67) and diastolic BP (β = 1.08 mm Hg, 95% CI: 0.66, 1.50) via BMI. The total effect of CD on systolic BP was not significant (β = 0.01 mm Hg, 95% CI: −1.36, 1.38), because the negative direct effect of CD on systolic BP cancelled out the positive indirect effect via BMI. A similar pattern was observed for diastolic BP (β = 0.70 mm Hg, 95% CI: −0.16, 1.55) (Table 2). DISCUSSION In this study, we followed up a population-based sample of 2,020 persons from birth to adulthood, and after adjustment for confounders, we found that the odds of adult hypertension later in life were 49% higher among persons born via CD. This association was at least partially mediated by adult BMI. Given the young age of these adults, diastolic BP could be more affected than systolic BP. While peripheral vascular resistance appears earlier and influences diastolic BP, altered vascular walls appear later and influence systolic BP (26). Young adults tend not to have vascular damage but to have increased adrenergic tone—and obesity is often associated with increased adrenergic discharge (27). Until recently, the consequences of surgical delivery have been evaluated according to its immediate association with maternal and newborn health. The hypothesis that it could also be linked to risk of future noncommunicable diseases introduces a new spectrum of outcomes to be considered. This is important because rapidly developing nations, such as Brazil, are among those with higher CD rates and an increasing burden of noncommunicable diseases. Our study adds to the evidence that CD may be associated with hypertension in adult life. Our group was one of the first to demonstrate an association between CD and obesity in adulthood (28) and in school-aged children from different socioeconomic backgrounds (5). We also found that CD was associated with increased peripheral and central adiposity (29) but not with other adult metabolic risk factors (30). A proposed mechanism by which CD may cause obesity is changes in the gut microbiota. Persons born via CD have been found to have a higher ratio of Firmicutes to Bacterioidetes compared with those delivered vaginally; predominance of Firmicutes in relation to Bacterioidetes in the gut microbiota is associated with increased fat absorption and obesity (31). It is possible that the mechanism hypothesized to cause obesity among persons born via CD leads afterward to hypertension, since it has long been known that obesity is one of the risk factors for hypertension. In our study, the association between CD and hypertension was partially mediated through BMI, supporting this hypothesis of a shared pathway. Studies confirming an association between gut microbiota and hypertension have already been published (32–34). We found a direct effect point estimate higher than that for the indirect effect, although nonsignificant. This does not eliminate the possibility of a nonmediated association. In fact, other pathways could have contributed to the association. Epigenetic changes have been associated with CD (35, 36). On the other hand, studies have demonstrated the importance of epigenetic regulation in the pathophysiology of hypertension (37). We did not identify any studies which have tested associations between altered gene expression linked to type of delivery and BP. In one previous study, Horta et al. (10) demonstrated an association between birth via CD and systolic BP in adulthood. There is evidence that a number of other maternal factors are associated with increasing BP in the offspring, including maternal hypertension (38), caffeine (39), and high salt intake during pregnancy (40). Among these factors, only maternal hypertension could be considered a confounder in a study like ours, because it is the only one linked to type of delivery. We controlled for maternal hypertension in our models. The use of a causal directed acyclic graph to understand and illustrate the relationships between variables allowed us to more clearly identify confounders and mediators. Common causes of overadjustment include controlling for variables that lie within the pathway that links the exposure to the outcome. In our mediation analysis, we adjusted for 3 sets of confounders: those that act in the association between CD and hypertension; those that act in the association between CD and adult BMI; and those that act in the association between adult BMI and hypertension. While it is expected that the indirect and direct effects would add up to the total effect, that was not the case in this study. This can be explained by noncollapsibility of the odds ratio, as demonstrated by Greenland and Morgenstern (41). One strength of this study is that it was conducted in a country with a high CD rate. Another is the estimation of total, direct, and indirect effects using the counterfactual definition, allowing for exposure-mediator interaction. However, some limitations must be highlighted. Exposure assessment (type of delivery) was performed immediately after birth, but BP assessed at one point in time is subject to measurement error. However, this would probably have resulted in misclassification that was nondifferential by exposure group. A second limitation was loss to follow-up, which produced a healthier sample than the original one. We minimized this limitation by performing inverse probability weighting. Another limitation is that we measured the mediator, the outcome, and the mediator-outcome confounders at the same time. Establishing temporal sequences becomes more difficult in such a situation. Our results were also vulnerable to hidden-variable bias. We were not able to perform weighted mediation analysis because the current version of “paramed” does not allow the use of inverse probability weights. However, differences between weighted and unweighted estimates were mostly small. Unmeasured mediator-outcome confounding could have persisted, even after we searched actively for more possible confounders. Finally, this cohort lacked some information that would be worth analyzing, such as indication for CD and maternal ethnicity. In conclusion, our results suggest that CD is associated with an increased risk of adult hypertension. There is growing evidence pointing in this direction in basic and epidemiologic research. We also found that this association is possibly mediated by BMI. This issue must be further investigated, since it has implications for countries that are already struggling with the burden of noncommunicable diseases and where high CD rates are found. Our findings contribute to the discussion on indications for elective CD. If the association between CD and hypertension mediated by BMI is causal, a decline in the CD rate could affect the long-term prevalence of both obesity and hypertension. Among the various prevention measures against noncommunicable diseases, which need the collaboration of various sectors of society, this one would depend greatly on the health sector. ACKNOWLEDGMENTS Author affiliations: Department of Pediatrics, Faculty of Medicine, University of São Paulo, São Paulo, Brazil (Alexandre Archanjo Ferraro, Maria Teresa Bechere Fernandes); Department of Pediatrics, Faculty of Medicine of Ribeirão Preto, University of São Paulo, Ribeirão Preto, Brazil (Marco Antônio Barbieri, Viviane Cunha Cardoso, Heloisa Bettiol); Department of Public Health, University of Maranhão, São Luis, Brazil (Antonio Augusto Moura da Silva); Department of Pediatrics, Faculty of Medicine, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil (Marcelo Zubaran Goldani); and Hubert Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, Georgia (Dr. Aryeh David Stein). This work was funded by the São Paulo State Research Foundation and by the Brazilian National Research Council. Conflict of interest: none declared. Abbreviations BMI body mass index BP blood pressure CD cesarean delivery CI confidence interval SD standard deviation REFERENCES 1 Victora CG , Aquino EM , do Carmo Leal M , et al. Maternal and child health in Brazil: progress and challenges . Lancet . 2011 ; 377 ( 9780 ): 1863 – 1876 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Betran AP , Torloni MR , Zhang J , et al. What is the optimal rate of caesarean section at population level? A systematic review of ecologic studies . Reprod Health . 2015 ; 12 : Article 57 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Kuhle S , Tong OS , Woolcott CG . Association between caesarean section and childhood obesity: a systematic review and meta-analysis . Obes Rev . 2015 ; 16 ( 4 ): 295 – 303 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Li HT , Zhou YB , Liu JM . The impact of cesarean section on offspring overweight and obesity: a systematic review and meta-analysis . Int J Obes (Lond) . 2013 ; 37 ( 7 ): 893 – 899 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Goldani MZ , Barbieri MA , da Silva AA , et al. Cesarean section and increased body mass index in school children: two cohort studies from distinct socioeconomic background areas in Brazil . Nutr J . 2013 ; 12 : Article 104 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Mamun AA , Sutharsan R , O’Callaghan M , et al. Cesarean delivery and the long-term risk of offspring obesity . Obstet Gynecol . 2013 ; 122 ( 6 ): 1176 – 1183 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Pei Z , Heinrich J , Fuertes E , et al. Cesarean delivery and risk of childhood obesity . J Pediatr . 2014 ; 164 ( 5 ): 1068 – 1073.e2 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Thavagnanam S , Fleming J , Bromley A , et al. A meta-analysis of the association between Caesarean section and childhood asthma . Clin Exp Allergy . 2008 ; 38 ( 4 ): 629 – 633 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Cardwell CR , Stene LC , Joner G , et al. Cesarean section is associated with an increased risk of childhood-onset type 1 diabetes mellitus: a meta-analysis of observational studies . Diabetologia . 2008 ; 51 ( 5 ): 726 – 735 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Horta BL , Gigante DP , Lima RC , et al. Birth by caesarean section and prevalence of risk factors for non-communicable diseases in young adults: a birth cohort study . PLoS One . 2013 ; 8 ( 9 ): e74301 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Haeffner LS , Barbieri MA , Rona RJ , et al. The relative strength of weight and length at birth in contrast to social factors as determinants of height at 18 years in Brazil . Ann Hum Biol . 2002 ; 29 ( 6 ): 627 – 640 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Cardoso VC , Simões VM , Barbieri MA , et al. Profile of three Brazilian birth cohort studies in Ribeirão Preto, SP and São Luís, MA . Braz J Med Biol Res . 2007 ; 40 ( 9 ): 1165 – 1176 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Olsen J , Frische G . Social differences in reproductive health. A study on birth weight, stillbirths and congenital malformations in Denmark . Scand J Soc Med . 1993 ; 21 ( 2 ): 90 – 97 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Davies DP , Holding RE . Neonatometer: a new infant length measurer . Arch Dis Child . 1972 ; 47 ( 256 ): 938 – 940 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Pardini R , Matsudo SM , Araújo T , et al. Validation of the International Physical Activity Questionaire (IPAQ-version 6): pilot study in Brazilian young adults [in Portuguese]. Rev Bras Ciên Mov . 2001 ; 9 ( 3 ): 45 – 51 . WorldCat 16 Ribeiro AB , Cardoso MA . Development of a food frequency questionnaire as a tool for programs of chronic diseases prevention . Rev Nutr . 2002 ; 15 ( 2 ): 239 – 245 . Google Scholar Crossref Search ADS WorldCat 17 Molina MC , Bettiol H , Barbieri MA , et al. Food consumption by young adults living in Ribeirão Preto, SP, 2002/2004 . Braz J Med Biol Res . 2007 ; 40 ( 9 ): 1257 – 1266 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Hense HW , Koivisto AM , Kuulasmaa K , et al. Assessment of blood pressure measurement quality in the baseline surveys of the WHO MONICA project . J Hum Hypertens . 1995 ; 9 ( 12 ): 935 – 946 . Google Scholar PubMed WorldCat 19 Cameron N . BASIC programs for the assessment of skeletal maturity and the prediction of adult height using the Tanner-Whitehouse method . Ann Hum Biol . 1984 ; 11 ( 3 ): 261 – 264 . Google Scholar Crossref Search ADS PubMed WorldCat 20 White IR , Royston P , Wood AM . Multiple imputation using chained equations: issues and guidance for practice . Stat Med . 2011 ; 30 ( 4 ): 377 – 399 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Wald DS , Law M , Morris JK , et al. Combination therapy versus monotherapy in reducing blood pressure: meta-analysis on 11,000 participants from 42 trials . Am J Med . 2009 ; 122 ( 3 ): 290 – 300 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Härkänen T , Kaikkonen R , Virtala E , et al. Inverse probability weighting and doubly robust methods in correcting the effects of non-response in the reimbursed medication and self-reported turnout estimates in the ATH survey . BMC Public Health . 2014 ; 14 : Article 1150 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Textor J , Hardt J , Knüppel S . DAGitty: a graphical tool for analyzing causal diagrams . Epidemiology . 2011 ; 22 ( 5 ): 745 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Pearl J . Causality: Models, Reasoning and Inference . 2nd ed . Cambridge, United Kingdom : Cambridge University Press ; 2009 . Google Preview WorldCat 25 Emsley R , Liu H . PARAMED: Stata Module to Perform Causal Mediation Analysis Using Parametric Regression Models. (Statistical Software Components S457581). Boston, MA: Department of Economics, Boston College; 2013 . https://ideas.repec.org/c/boc/bocode/s457581.html. Accessed April 3, 2019. 26 Franklin SS , Gustin W , Wong ND , et al. Hemodynamic patterns of age-related changes in blood pressure—the Framingham Heart Study . Circulation . 1997 ; 96 ( 1 ): 308 – 315 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Oparil S , Zaman MA , Calhoun DA . Pathogenesis of hypertension . Ann Intern Med . 2003 ; 139 ( 9 ): 761 – 776 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Goldani HA , Bettiol H , Barbieri MA , et al. Cesarean delivery is associated with an increased risk of obesity in adulthood in a Brazilian birth cohort study . Am J Clin Nutr . 2011 ; 93 ( 6 ): 1344 – 1347 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Mesquita DN , Barbieri MA , Goldani HA , et al. Cesarean section is associated with increased peripheral and central adiposity in young adulthood: cohort study . PLoS One . 2013 ; 8 ( 6 ): e66827 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Bernardi JR , Pinheiro TV , Mueller NT , et al. Cesarean delivery and metabolic risk factors in young adults: a Brazilian birth cohort study . Am J Clin Nutr . 2015 ; 102 ( 2 ): 295 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Clarke SF , Murphy EF , Nilaweera K , et al. The gut microbiota and its relationship to diet and obesity: new insights . Gut Microbes . 2012 ; 3 ( 3 ): 186 – 202 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Jose PA , Raj D . Gut microbiota in hypertension . Curr Opin Nephrol Hypertens . 2015 ; 24 ( 5 ): 403 – 409 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Yang T , Santisteban MM , Rodriguez V , et al. Gut dysbiosis is linked to hypertension . Hypertension . 2015 ; 65 ( 6 ): 1331 – 1340 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Li J , Zhao F , Wang Y , et al. Gut microbiota dysbiosis contributes to the development of hypertension . Microbiome . 2017 ; 5 : Article 14 . Google Scholar Crossref Search ADS WorldCat 35 Schlinzig T , Johansson S , Gunnar A , et al. Epigenetic modulation at birth altered DNA-methylation in white blood cells after cesarean section . Acta Paediatr . 2009 ; 98 ( 7 ): 1096 – 1099 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Franz MB , Poterauer M , Elhenicky M , et al. Global and single gene DNA methylation in umbilical cord blood cells 2 after elective caesarean: a pilot study . Eur J Obstet Gynecol Reprod Biol . 2014 ; 179 : 121 – 124 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Arnett DK , Claas SA . Omics of blood pressure and hypertension . Circ Res . 2018 ; 122 ( 10 ): 1409 – 1419 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Kajantie E , Eriksson JG , Osmond C , et al. Pre-eclampsia is associated with increased risk of stroke in the adult offspring: the Helsinki Birth Cohort Study . Stroke . 2009 ; 40 ( 4 ): 1176 – 1180 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Serapiao-Moraes DF , Souza-Mello V , Aguila MB , et al. Maternal caffeine administration leads to adverse effects on adult mice offspring . Eur J Nutr . 2013 ; 52 ( 8 ): 1891 – 1900 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Reynolds CM , Vickers MH , Harrison CJ , et al. High fat and/or high salt intake during pregnancy alters maternal meta-inflammation and offspring growth and metabolic profiles . Physiol Rep . 2014 ; 2 ( 8 ):pii:e12110. WorldCat 41 Greenland S , Morgenstern H . Confounding in health research . Annu Rev Public Health . 2001 ; 22 : 189 – 212 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2019. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Cesarean Delivery and Hypertension in Early Adulthood
JF - American Journal of Epidemiology
DO - 10.1093/aje/kwz096
DA - 2019-07-01
UR - https://www.deepdyve.com/lp/oxford-university-press/cesarean-delivery-and-hypertension-in-early-adulthood-V5svI2P0gT
SP - 1296
VL - 188
IS - 7
DP - DeepDyve
ER -