To Örebro University

Background: Maternal type 1 diabetes (T1D) has been linked to preterm birth and other adverse pregnancy outcomes. How these risks vary with glycated hemoglobin (or hemoglobin A(1c) [HbA(1c)]) levels is unclear.

Objective: To examine preterm birth risk according to periconceptional HbA(1c) levels in women with T1D.

Design: Population-based cohort study.

Setting: Sweden, 2003 to 2014.

Patients: 2474 singletons born to women with T1D and 1 165 216 reference infants born to women without diabetes.

Measurements: Risk for preterm birth (< 37 gestational weeks). Secondary outcomes were neonatal death, large for gestational age, macrosomia, infant birth injury, hypoglycemia, respiratory distress, 5-minute Apgar score less than 7, and stillbirth. Results: Preterm birth occurred in 552 (22.3%) of 2474 infants born to mothers with T1D versus 54 287 (4.7%) in 1 165 216 infants born to mothers without diabetes. The incidence of preterm birth was 13.2% in women with a periconceptional HbA(1c) level below 6.5% (adjusted risk ratio [aRR] vs. women without T1D, 2.83 [95% CI, 2.28 to 3.52]), 20.6% in those with a level from 6.5% to less than 7.8% (aRR, 4.22 [CI, 3.74 to 4.75]), 28.3% in those with a level from 7.8% to less than 9.1% (aRR, 5.56 [CI, 4.84 to 6.38]), and 37.5% in those with a level of 9.1% or higher (aRR, 6.91 [CI, 5.85 to 8.17]). The corresponding aRRs for medically indicated preterm birth (n = 320) were 5.26 (CI, 3.83 to 7.22), 7.42 (CI, 6.21 to 8.86), 11.75 (CI, 9.72 to 14.20), and 17.51 (CI, 14.14 to 21.69), respectively. The corresponding aRRs for spontaneous preterm birth (n = 223) were 1.81 (CI, 1.31 to 2.52), 2.86 (CI, 2.38 to 3.44), 2.88 (CI, 2.23 to 3.71), and 2.80 (CI, 1.94 to 4.03), respectively. Increasing HbA(1c) levels were associated with the study's secondary outcomes: large for gestational age, hypoglycemia, respiratory distress, low Apgar score, neonatal death, and stillbirth.

Limitation: Because HbA(1c) levels were registered annually at routine visits, they were not available for all pregnant women with T1D.

Conclusion: The risk for preterm birth was strongly linked to periconceptional HbA(1c) levels. Women with HbA(1c) levels consistent with recommended target levels also were at increased risk. Primary Funding Source: Swedish Diabetes Foundation.

Education, income, and occupation are factors known to affect health and disease. In this review we describe the Swedish Longitudinal Integrated Database for Health Insurance and Labour Market Studies (LISA, Longitudinell Integrationsdatabas for Sjukforsakrings- och Arbetsmarknadsstudier). LISA covers the adult Swedish population aged16years registered on December 31 each year since 1990 (since 2010 individuals aged15years). The database was launched in response to rising levels of sick leave in the country. Participation in Swedish government-administered registers such as LISA is compulsory, and hence selection bias is minimized. The LISA database allows researchers to identify individuals who do not work because of injury, disease, or rehabilitation. It contains data on sick leave and disability pension based on calendar year. LISA also includes information on unemployment benefits, disposable income, social welfare payments, civil status, and migration. During 2000-2017, an average of 97,000 individuals immigrated to Sweden each year. This corresponds to about 1% of the Swedish population (10 million people in 2017). Data on occupation have a completeness of 95%. Income data consist primarily of income from employment, capital, and allowances, including parental allowance. In Sweden, work force participation is around 80% (2017: overall: 79.1%; men 80.3% and women 77.9%). Education data are available in>98% of all individuals aged 25-64years, with an estimated accuracy for highest attained level of education of 85%. Some information on civil status, income, education, and employment before 1990 can be obtained through the Population and Housing Census data (FoB, Folk- och bostadsrakningen).

Background: Childhood-onset inflammatory bowel disease (IBD) might negatively impact academic school performance. We conducted a nationwide study to examine the association between childhood-onset IBD and school results.

Methods: Our study population was selected from Swedish health registers. In the National Patient Register (1990 to 2013), we identified 2827 children with IBD: Crohn's disease (CD), n = 1207, and ulcerative colitis (UC), n = 1370. Patients were matched with 10 reference individuals by age, sex, birth year, and place of residence (n = 28,235). Final compulsory school grades (0 to 320 grade points) and qualification for high school (yes or no) were obtained through the National School Register. Regression models controlling for socioeconomic factors were used to analyze the association of IBD with school performance.

Results: Children with IBD had a lower final grade point average (adjusted mean grade difference [AMGD] -4.9, 95% confidence interval [CI] -7.1 to -2.6) but not a significantly higher risk to not qualify for high school (odds ratio [OR] 1.14, CI 0.99-1.31). The results were similar in children with UC (AMGD -5.5, CI -8.7 to -2.3) and CD (AMGD -4.7, CI -8.2 to -1.2). Underperformance was more common in subsets of IBD children characterized by markers associated with long-standing active disease (eg, >30 inpatient days [AMGD-18.1, CI -25.8 to -10.4]).

Conclusion: Most children with IBD achieve comparable results in the final year of compulsory school as their healthy peers. Care should be improved for the subgroup of children for which IBD has a stronger negative impact on school performance.

OBJECTIVES: To assess whether male elite football players are at increased risk of alcohol related disorders compared with men from the general population, and whether such an increased risk would vary on the basis of calendar year of the first playing season in the top tier of competition, age, career length, and goal scoring abilities.

DESIGN: Nationwide cohort study.

SETTING: Sweden, 1924-2020.

PARTICIPANTS: 6007 male football players who had played in the Swedish top division, Allsvenskan, from 1924 to 2019 and 56 168 men from the general population matched to players based on age and region of residence.

MAIN OUTCOME MEASURES: Primary outcome was alcohol related disorders (diagnoses recorded in death certificates, during hospital admissions and outpatient visits, or use of prescription drugs for alcohol addiction); secondary outcome was disorders related to misuse of other drugs.

RESULTS: During follow-up up to 31 December 2020, 257 (4.3%) football players and 3528 (6.3%) men from the general population received diagnoses of alcohol related disorders. In analyses accounting for age, region of residence, and calendar time, risk of alcohol related disorders was lower among football players than among men from the general population (hazard ratio 0.71, 95% confidence interval 0.62 to 0.81). A reduced risk of alcohol related disorders was observed for football players who played their first season in the top tier in the early 1960s and later, while no significant difference versus men from the general population was seen in the risk for football players from earlier eras. The hazard ratio was lowest at around age 35 years, and then increased with age; at around age 75 years, football players had a higher risk of alcohol related disorders than men from the general population. No significant association was seen between goal scoring, number of games, and seasons played in the top tier and the risk of alcohol related disorders. Risk of disorders related to other drug misuse was significantly lower among football players than the general population (hazard ratio 0.22, 95% confidence interval 0.15 to 0.34).

CONCLUSIONS: In this nationwide cohort study, male football players who had played in the Swedish top tier of competition had a significantly lower risk of alcohol related disorders than men from the general population.