Epigenome-wide association study of DNA methylation in maternal blood leukocytes with BMI in pregnancy and gestational weight gain.

OBJECTIVES: We aimed to discover CpG sites with differential DNA methylation in peripheral blood leukocytes associated with body mass index (BMI) in pregnancy and gestational weight gain (GWG) in women of European and South Asian ancestry. Furthermore, we aimed to investigate how the identified sites were associated with methylation quantitative trait loci, gene ontology, and cardiometabolic parameters. METHODS: In the Epigenetics in pregnancy (EPIPREG) sample we quantified maternal DNA methylation in peripheral blood leukocytes in gestational week 28 with Illumina's MethylationEPIC BeadChip. In women with European (n = 303) and South Asian (n = 164) ancestry, we performed an epigenome-wide association study of BMI in gestational week 28 and GWG between gestational weeks 15 and 28 using a meta-analysis approach. Replication was performed in the Norwegian Mother, Father, and Child Cohort Study, the Study of Assisted Reproductive Technologies (MoBa-START) (n = 877, mainly European/Norwegian). RESULTS: We identified one CpG site significantly associated with GWG (p 5.8 × 10-8) and five CpG sites associated with BMI at gestational week 28 (p from 4.0 × 10-8 to 2.1 × 10-10). Of these, we were able to replicate three in MoBa-START; cg02786370, cg19758958 and cg10472537. Two sites are located in genes previously associated with blood pressure and BMI. DNA methylation at the three replicated CpG sites were associated with levels of blood pressure, lipids and glucose in EPIPREG (p from 1.2 × 10-8 to 0.04). CONCLUSIONS: We identified five CpG sites associated with BMI at gestational week 28, and one with GWG. Three of the sites were replicated in an independent cohort. Several genetic variants were associated with DNA methylation at cg02786379 and cg16733643 suggesting a genetic component influencing differential methylation. The identified CpG sites were associated with cardiometabolic traits. GOV REGISTRATION NO: Not applicable.

Variation in the achievement of HbA1c , blood pressure and LDL cholesterol targets in type 2 diabetes in general practice and characteristics associated with risk factor control.

AIMS: To identify population, general practitioner, and practice characteristics associated with the achievement of HbA1c , blood pressure and LDL cholesterol targets, and to describe variation in the achievement of risk factor control. METHODS: We conducted a cross-sectional survey of 9342 people with type 2 diabetes, 281 general practitioners and 77 general practices in Norway. Missing values (7.4%) were imputed using multiple imputation by chained equations. We used three-level logistic regression with the achievement of HbA1c , blood pressure and LDL cholesterol targets as dependent variables, and factors related to population, general practitioners, and practices as independent variables. RESULTS: Treatment targets were achieved for HbA1c in 64%, blood pressure in 50%, and LDL cholesterol in 52% of people with type 2 diabetes, and 17% met all three targets. There was substantial heterogeneity in target achievement among general practitioners and among practices; the estimated proportion of a GPs diabetes population at target was 55-73% (10-90 percentiles) for HbA1c , 36-63% for blood pressure, and 47-57% for LDL cholesterol targets. The models explained 11%, 5% and 14%, respectively, of the total variation in the achievement of HbA1c , blood pressure and LDL cholesterol targets. Use among general practitioners of a structured diabetes form was associated with 23% higher odds of achieving the HbA1c target (odds ratio 1.23, 95% confidence interval (CI) 1.02-1.47) and 17% higher odds of achieving the LDL cholesterol target (odds ratio 1.17, 95% CI 1.01-1.35). CONCLUSIONS: Clinical diabetes management is difficult, and few people meet all three risk factor control targets. The proportion of people reaching target varied among general practitioners and practices. Several population, general practitioner and practice characteristics only explained a small part of the total variation. The use of a structured diabetes form is recommended.

Phenotypic and genotypic differences between Indian and Scandinavian women with gestational diabetes mellitus.

OBJECTIVE: Gestational diabetes mellitus (GDM) is a transient form of diabetes characterized by impaired insulin secretion and action during pregnancy. Population-based differences in prevalence exist which could be explained by phenotypic and genetic differences. The aim of this study was to examine these differences in pregnant women from Punjab, India and Scandinavia. METHODS: Eighty-five GDM/T2D loci in European and/or Indian populations from previous studies were assessed for association with GDM based on Swedish GDM criteria in 4018 Punjabi Indian and 507 Swedish pregnant women. Selected loci were replicated in Scandinavian cohorts, Radiel (N = 398, Finnish) and STORK/STORK-G (N = 780, Norwegian). RESULTS: Punjabi Indian women had higher GDM prevalence, lower insulin secretion and better insulin sensitivity than Swedish women. There were significant frequency differences of GDM/T2D risk alleles between both populations. rs7178572 at HMG20A, previously associated with GDM in South Indian and European women, was replicated in North Indian women. The T2D risk SNP rs11605924 in the CRY2 gene was associated with increased GDM risk in Scandinavian but decreased GDM risk in Punjabi Indian women. No other overlap was seen between GDM loci in both populations. CONCLUSIONS: Gestational diabetes mellitus is more common in Indian than Swedish women, which partially can be attributed to differences in insulin secretion and action. There was marked heterogeneity in the GDM phenotypes between the populations which could only partially be explained by genetic differences.

Population, general practitioner and practice characteristics are associated with screening procedures for microvascular complications in Type 2 diabetes care in Norway.

AIMS: To assess population, general practitioner (GP) and practice characteristics associated with the performance of microvascular screening procedures and to propose strategies to improve Type 2 diabetes care. METHODS: A cross-sectional survey in Norway (281 GPs from 77 practices) identified 8246 people with a Type 2 diabetes duration of 1 year or more. We used multilevel regression models with either the recording of at least two of three recommended screening procedures (albuminuria, monofilament, eye examination) or each procedure separately as dependent variable (yes/no), and characteristics related to the person with diabetes, GP or practice as independent variables. RESULTS: The performance of recommended screening procedures was recorded in the following percentages: albuminuria 31.5%, monofilament 27.5% and eye examination 60.0%. There was substantial heterogeneity between practices, and between GPs within practices for all procedures. Compared with people aged 60-69 years, those aged < 50 years were less likely to have an albuminuria test performed [odds ratio (OR) 0.75, 95% CI 0.61 to 0.93] and eye examination (OR 0.79, 95% CI 0.66 to 0.95). People with macrovascular disease had fewer screening procedures recorded (OR 0.68, 95% CI 0.59 to 0.78). Use of an electronic diabetes form was associated with improved screening  (OR 2.65, 95% CI 1.86 to 3.78). GPs with high workload recorded fewer procedures (OR 0.59, 95% CI 0.39 to 0.90). CONCLUSIONS: Performance of screening procedures was suboptimal overall, and in people who should be prioritized. Performance varied substantially between GPs and practices. The use of a structured diabetes form should be mandatory.

Ethnic differences in postpartum weight retention: a Norwegian cohort study.

OBJECTIVE: To explore ethnic differences in weight retention 14 weeks postpartum. DESIGN: Population-based cohort study. SETTING: The STORK Groruddalen Study. POPULATION: A multi-ethnic cohort of healthy pregnant women attending primary antenatal care at three public Child Health Clinics, in Oslo, Norway (n = 642). METHODS: An explanatory linear regression was performed to model the relationship between ethnicity and postpartum weight retention. Forward selection of 12 explanatory factors was used to adjust for potential confounding factors, based on univariate analysis and adjusted R(2) . MAIN OUTCOME MEASURE: Postpartum weight retention. RESULTS: Unadjusted mean postpartum weight retention was 2.3 (4.9) kg for women from Western Europe and varied from 3.7 (3.5) to 6.3 (4.7) kg among the five ethnic minority groups. The proportion of women in the highest quintile (postpartum weight retention >8.5-24.4 kg) significantly differed by ethnicity (P < 0.01 for the proportion of women from South Asia, the Middle East and Africa compared with Western Europeans). Women from all ethnic minority groups had a higher relative increase in weight from pre-pregnancy to postpartum (P < 0.01) compared with Western Europeans. After adjustments for significant exposures, women from the Middle East retained 2.0 kg (95% CI: 1.0-3.0), South Asia 2.8 kg (91.9-3.6), and Africa 4.4 kg (3.1-5.8) more than Western Europeans (P < 0.01). CONCLUSIONS: Significantly more women with an ethnic origin from South Asia, the Middle East and Africa had high postpartum weight retention compared with Western European women.

Weight gain, total fat gain and regional fat gain during pregnancy and the association with gestational diabetes: a population-based cohort study.

OBJECTIVE: In a multi-ethnic population-based study, we investigate the change in indicators of adiposity (being weight gain and gain of total fat, truncal fat and mean skinfold thickness) from early pregnancy to 28 weeks of gestation overall and across ethnic groups, and explore the association between the change in indicators of adiposity and gestational diabetes (GDM). DESIGN: Weight, skinfold thickness and bioelectrical impedance analysis were performed twice in 728 pregnant women in gestational week 15 (visit 1) and week 28 (visit 2). GDM was defined by the modified International Association of Diabetes in Pregnancy Study Group (IADPSG) criteria (1-hour glucose not available). RESULTS: An increase in all indicators of adiposity gave increased odds ratios (OR) for GDM. After adjusting for pre-pregnant body mass index, a 0.14 kg per week (one standard deviation (s.d.)) increase in truncal fat gave an OR of 1.31 (95% CI 1.10-1.56), while a 0.21 kg per week (one s.d.) weight gain gave an OR of 1.23 (95% CI 1.04-1.46) for GDM. The ORs for the indicators of adiposity remained after additional adjustments for insulin resistance in early pregnancy. When combining the effects of an ethnic origin, 0.14 kg per week (one s.d.) truncal fat gain and 4.7 kg m(-2) (one s.d.) increased pre-pregnant BMI the OR for South Asians was 5.9 (3.5-10.0) versus 2.1 (1.6-2.8) for Europeans. CONCLUSION: Weight gain and gain of total fat mass, mean skinfold thickness and especially truncal fat were all positively associated with GDM. South Asians, in particular, should be encouraged to avoid an excessive weight gain during pregnancy to reduce risk of GDM.

Objectively recorded physical activity in early pregnancy: a multiethnic population-based study.

This study aimed to compare objectively recorded physical activity (PA) levels and walking steps among pregnant women. Cross-sectional data from a multiethnic cohort (n = 823) of pregnant women consisting of 44% from Western countries, 24% from South Asia, 14% from Middle East, and 18% from other countries. PA and steps were recorded by the activity monitor SenseWear™ Pro3 Armband. A total of 678 women were included in the analysis. Western women walked significantly more steps and had higher moderate-to-vigorous-intensity physical activity (MVPA) levels compared with South Asian women per weekday and weekend day. Interaction terms (P = 0.008) between ethnicity (Western vs South Asian) and parity, and education, respectively, were identified: having ≥ 1 children was positively associated with steps during weekends in South Asians in contrast to Western women. Having <12 years education was associated with more MVPA time among South Asians in contrast to Western women. South Asian women are prone to low levels of PA during pregnancy and South Asian women without children and with higher education may have an elevated risk for an inactive lifestyle during pregnancy.

Objectively recorded physical activity and the association with gestational diabetes.

The aim of this population-based study was to assess the association between objectively recorded physical activity (PA) in early gestation and gestational diabetes mellitus (GDM) identified at 28 weeks of gestation in a multi-ethnic cohort of healthy pregnant women in Oslo, Norway. In total, 759 women were included. In early gestation (<20 weeks), light-, moderate-, and vigorous-intensity PA and number of steps were objectively recorded (SenseWear™ Armband Pro3), and self-reported PA, demographics, and anthropometrics were collected. The 75-g oral glucose tolerance test was performed at 28 weeks of gestation. Women with GDM had fewer objectively recorded steps (mean 7964 steps/day vs 8879 steps/day, P < 0.001) and minutes of moderate-to-vigorous-intensity PA (median 62 min/day vs 75 min/day, P = 0.004) in early gestation than women without GDM. Additionally, 30% of women with GDM compared with 44% (P < 0.001) of women without GDM self-reported regular PA before pregnancy. The significant inverse association between objectively recorded steps per day in early gestation and GDM persisted after adjustment for ethnic origin, weeks of gestation, age, parity, pre-pregnancy BMI, early life socioeconomic position, and self-reported regular PA before pregnancy. The adjusted odds ratio for GDM decreased 19% per standard deviation (3159 steps) increase in objectively recorded steps per day (P = 0.039). Daily life PA in early gestation measured as steps/day was associated with lower risk of GDM.

The prevalence of urinary incontinence in pregnancy among a multi-ethnic population resident in Norway.

OBJECTIVES: To investigate prevalence of urinary incontinence (UI) in a multi-ethnic population of pregnant women, and to analyse for possible associations of the known risk factors for UI in such a population. DESIGN: Population-based cross-sectional study. SETTING: All pregnant women in three administrative city districts attending the Child Health Clinics. POPULATION AND SAMPLE: Out of 823 women identified in the [corrected] first trimester, 772 (94%) [corrected] agreed to participate in the study at 28 weeks of gestation. Inclusion criteria were: healthy women at 20 weeks of gestation or less and able to communicate in Norwegian, Arabic, English, Sorani, Somali, Tamil, Turkish, Urdu or Vietnamese. METHODS: Differences between ethnic groups were tested by simple descriptive statistics. Associations were estimated by logistic regression analysis and presented as crude (cOR) and adjusted (aOR) odds ratios. MAIN OUTCOME MEASURES: Prevalence of UI as ascertained using the International Consultation on Incontinence Questionnaire-urinary incontinence-short form. RESULTS: Prevalence rates of UI at 28 weeks of gestation were 26% for women of African origin, 36% for women of Middle Eastern origin, 40% for women of East Asian origin, 43% for women of South Asian origin and 45% for women of European/North American origin. The difference was significant between women of African and European/North American origins (P = 0.011) and between women of African and South Asian origins (P = 0.035). Age (aOR 1.05; 95% CI 1.01-1.09) and parity (aOR 2.34; 95% CI 1.66-3.28) were positively associated with the prevalence of UI in pregnancy. Women of African origin had significantly reduced odds for UI (aOR 0.42; 95% CI 0.20-0.87). East Asian and African women reported the highest perceived impact of UI in pregnancy. CONCLUSIONS: A high prevalence of UI was found in a multi-ethnic pregnant population.

Targeting physical activity in a low socioeconomic status population: observations from the Norwegian 'Romsas in Motion' study.

OBJECTIVE: To summarise the main results of a community-based study on physical activity promotion - "Romsås in Motion". METHODS: The study assessed changes in physical activity, body mass and psychosocial mediators of physical activity from a pseudo-experimental cohort study involving two districts with low socioeconomic status in Oslo, Norway. In 2000, baseline investigation included 2950 30-67-year-olds - 48% of those invited. At follow-up in 2003 we measured 1776 subjects (67% of those eligible). A set of theoretically informed strategies targeting individuals, groups and the environment were implemented, tailored towards groups with different psychosocial readiness for change. Net changes (the difference between changes in the intervention and control districts) and results of mediation analyses related to the effect of the intervention are reported. RESULTS: The increase in physical activity measured by two questionnaires was 9.5% (p = 0.008) and 8.1% (p = 0.02), respectively. The proportion who increased their body mass was reduced by 50% compared with the control district. Participation in walking and aerobic exercise groups, having seen the "Walk the stairs" poster and used the walk path, were particularly effective intervention components. The most promising psychosocial mediators of forward transition in stages of change were physical activity identity, perceived control, support from friends and family, and self-efficacy when facing psychological barriers. CONCLUSION: Through a theoretically informed, low-cost, population-based intervention programme an increase in physical activity levels and a reduced weight gain were observed. Mediation findings regarding forward transition in stages of change enhance our understanding of psychosocial mechanisms of behaviour change, and may prove helpful in guiding implementation and evaluation of future interventions.

Differences in cardiovascular disease mortality and major risk factors between districts in Oslo. An ecological analysis.

BACKGROUND: The level of mortality from cardiovascular disease (CVD), coronary heart disease (CHD) and from all causes varies considerably within Oslo. The purpose of this study was to examine these differences according to cardiovascular risk factors and socioeconomic variables at the district level. METHODS: Total mortality rates and cardiovascular mortality rates for subjects aged 45-74 years in 1991-1995, and their relationship to cardiovascular risk factors and socioeconomic indicators in the 25 districts of Oslo were studied. Cardiovascular risk factors were based on data from 40 year olds in 1985-1988. The following variables were used as independent variables in the regression analyses to explain differences between the districts: daily smoking, cholesterol level, systolic blood pressure, education and income. RESULTS: Mortality rates were strongly related to cardiovascular risk factors and to socio-economic indicators, with correlation coefficients (Pearson) of 0.74 for smoking and CVD mortality, and -0.78 for high income and CHD mortality. Smoking explained 70% of the differences in mortality from all causes for men and 46% for women, and 61% and 49% of the differences in CVD mortality for men and women respectively. Diastolic blood pressure and total cholesterol were closely related to socioeconomic indicators and to smoking, but the relative strength of the cardiovascular risk factors in the multivariate analyses differed for males and females. CONCLUSION: diovascular risk factors and socioeconomic indicators at the population level were strongly related to mortality, and explained a large proportion of the differences in mortality between different districts of Oslo in the 1990s.

Ethnicity and sex are strong determinants of diabetes in an urban Western society: implications for prevention.

AIMS/HYPOTHESIS: This study was conducted to investigate the prevalence of diabetes and its association with ethnicity and sex, to identify subgroups at special risk. METHODS: We performed a population-based cross-sectional survey of 30- to 67-year-olds in an area of Oslo with low socio-economic status, and collected data using questionnaires, physical examinations and serum analyses for the 2,513 participants (attendance rate 49.3%). RESULTS: In the age group 30-59 years, mean BMI was 28.5 (95% CI: 27.5-29.6) for South Asian women, 26.1 (25.9-26.4) for Western women, 26.7 (26.1-27.4) for South Asian men and 27.2 (26.9-27.5) for Western men. The diabetes prevalence rates were 27.5% (18.1-36.9) for South Asian women, 2.9% (1.9-3.4) for Western women, 14.3% (8.0-20.7) for South Asian men and 5.9% (4.2-7.5) for Western men. The age-adjusted odds ratio (OR) for diabetes for women vs men was 1.9 (0.9-4.1) for South Asians, and 0.4 (0.3-0.6) for the Western population (p<0.001). The age-adjusted OR for diabetes for South Asians vs Westerners was 11.0 (5.8-21.1) for women and 3.0 (1.6-5.4) for men, and after adjustment for WHR the ORs were 7.7 (3.9-15.3) for women and 2.6 (1.4-4.9) for men. After additional adjustments for physical activity, education, body height and fertility for women, the OR was 6.0 (2.3-15.4) for women and 1.9 (0.9-4.0) for men. CONCLUSIONS/INTERPRETATION: The alarmingly high prevalence of diabetes among South Asian women in Norway needs further investigation, as it has considerable public health implications. Ethnic differences in OR for diabetes persisted after adjustment for age, adiposity, physical activity and education. These differences were still present for women after additional adjustment for body height and fertility.

[Cardiovascular risk factors in Oslo, Sogn and Fjordane and Finnmark. Study of 40-year old subjects in Oslo 1981-88]. / Risikofaktorer for hjerte- og karsykdom i Oslo, Sogn og Fjordane og Finnmark. 40-åringsundersøkelsen i Oslo 1981-88.

Cardiovascular risk factor levels among 40-year olds in Oslo were studied from the early 1970s until the late 1980s. Comparison has been made with similar data for the same period from two other regions in the country. The serum cholesterol level in men decreased from 6.3 mmol/l in 1972-73 to 5.9 mmol/l in 1985-88 and the proportion of men who smoked on a daily basis decreased from 51.8% to 40.8%. The proportion of female smokers increased from 37.4% in 1981-84 to 39.5% in 1985-88. There was a decrease in systolic blood pressure among men from 1972-73 to 1985-88. From 1981-84 to 1985-88 there was a minor increase for both sexes in the proportion of physically inactive for both sexes or those with a body mass index > 30. Prior to 1980 cardiovascular risk factors were most favourable in Sogn og Fjordane, and least favourable in Finnmark, with men from Oslo positioned somewhere in between these counties. With the exception of women's smoking habits and, for both sexes, the degree of physical activity, people living in Oslo in 1985-88 had a more favourable risk factor status than those in Sogn og Fjordane in 1985 and in Finnmark in 1987. Despite a degree of uncertainty because of methodological issues, the results indicate a reduction in risk factors of cardiovascular disease among men in Oslo during the 15 years covered by the study.

[Regional differences in risk factors in Oslo. Smoking, exercise, body mass index, blood lipids and blood pressure in 40-year old subjects 1985-88]. / Regionale ulikheter i sykdomsrisiko i Oslo. Røyke- og mosjonsvaner, kroppsmasseindeks, blodlipidnivå og blodtrykk blant 40-åringer 1985-88.

The general mortality rate in Oslo (capital city of Norway) is high compared with the rest of the country, and varies considerably from area to area within the city. This study highlights the differences in cardiovascular risk factors among 40 year-olds in different areas of Oslo between 1985 and 88. A total of 14,220 persons were included in the study. The city was divided into four geographical areas with distinct differences in socio-economic conditions. For both sexes all risk factors were shown to be more favourable in the more affluent areas than in the poorer area of the inner city. Also for both sexes the proportion of persons who smoked on a daily basis varied from 29% in the "best" area to 49% in the "worst" area of the city. The differences in serum cholesterol in the different areas were less pronounced than for the other risk factors. The greatest differences in risk factors were observed among women. The proportion of physically inactive women was 27% in the poorer area compared with 17% in the more affluent part of the city, and the body mass index was above 30 for 8.8% and 2.8% of women respectively. Similarly, the proportion of women with serum cholesterol > 7 mmol/l, triglycerides > or = 2.5 mmol/l, systolic blood pressure > or = 140 mm or diastolic blood pressure > or = 90 mm was almost twice as high for those living in poor socioeconomic conditions than those who were in a better position.