Log-linear Binomial Model Research Articles

Mother’s dietary quality during pregnancy and offspring’s dietary quality in adolescence: Follow-up from a national birth cohort study of 19,582 mother–offspring pairs

BackgroundThe Developmental Origins of Health and Disease (DOHaD) hypothesis postulates that exposures during early life, such as maternal dietary intake during pregnancy, may have a lifelong impact on the individual’s susceptibility to diseases. The individual’s own lifestyle habits are obviously an additional factor, but we have only limited knowledge regarding how it may interact with prenatal exposures in determining later disease. To gain further insight into these potentially complex relationships, we examined the longitudinal association between maternal diet quality during pregnancy and diet quality in early adolescence in a contemporary cohort.Methods and findingsFrom 1996 to 2003, the Danish National Birth Cohort (DNBC) was established. Women from across the country were enrolled, and dietary intake in midpregnancy was assessed concurrently with a 360-item food frequency questionnaire (FFQ) (https://www.dnbc.dk/-/media/arkiv/projekt-sites/dnbc/kodeboeger/dnbc-food-frequency-questionnaire/dnbc-food-frequency-questionnaire-pdf.pdf?la=en). During 2013–2018, dietary intake was assessed at age 14 years with a 150-item FFQ (https://www.dnbc.dk/-/media/arkiv/projekt-sites/dnbc/kodeboeger/ffq-14/dnbc-ffq-14-english-translation.pdf?la=en) in the DNBC children. Among the 19,582 mother–offspring pairs included in the analyses, the mean age (±standard deviation [SD]) was 30.7 (±4.1) years and 14.0 (±0.0) years for mothers and offspring, respectively. The majority of both mothers (67%) and offspring (76%) were classified as normal weight. For both questionnaires, a Healthy Eating Index (HEI) was developed as an indicator for diet quality based on current Danish Food-Based Dietary Guidelines (FBDG) including eight components: fruits and vegetables, fish, dietary fibres, red meat, saturated fatty acids (SFAs), sodium, sugar-sweetened beverages (SSBs), and added sugar. The HEI score was divided into quartiles; individuals in the highest quartile represented those with the most optimal diet. The maternal HEI score was correlated positively with offspring HEI score (Pearson r = 0.22, p < 0.001). A log-linear binomial model was used to estimate the relative risk of the offspring being in the highest quartile of HEI at age 14 years if the mother was ranked in quartile 4 during pregnancy. Results showed that offspring born to mothers who were in the highest HEI quartile during pregnancy were more likely themselves to be located in the highest HEI quartile at age 14 years (risk ratio [RR]: 2.1, 95% confidence interval [CI]: 2.0, 2.3, p < 0.001). Adjusting for maternal prepregnancy body mass index (BMI), parity, education, alcohol intake, physical activity, smoking, and breastfeeding, as well as offspring total energy intake and sex, did not influence the effect estimates. The limitations of our study include that some attrition bias towards more healthy participants was observed when comparing participants with nonparticipants. Bias in the FFQ method may also have resulted in underrepresentation of adolescents with poorer diet quality.ConclusionsIn this study using data from a large national birth cohort, we observed that maternal diet quality during pregnancy was associated with diet quality of the offspring at age 14 years. These findings indicate the importance of separating early dietary exposures from later dietary exposures when studying dietary aetiologies of diseases postulated to have developmental origins such as, for instance, obesity or asthma in observational settings.

Association between maternal diet quality and offspring diet quality assessed at age 14 years: longitudinal study in a large contemporary cohort

BackgroundEmerging evidence suggests that exposures in fetal life may have an impact many years later on the individual's susceptibility to diseases. Study of diet–disease associations over long time-spans is a major methodological challenge. We examined whether overall diet quality during pregnancy was associated with adolescent diet quality 15 years later in a prospective cohort representing a typical high-income population. MethodsDietary intake assessed with a 360-item food frequency questionnaire during mid-pregnancy (1996–2003) was matched with offspring dietary intake assessed with a 150-item food frequency questionnaire at age 14 years in 19 620 mother–offspring pairs from the Danish National Birth Cohort (a large cohort of pregnant women with long-term follow-up of the offspring). Only children still in the cohort, living with mother, father, or both parents, and with valid parental personal identification number were invited to complete the food frequency questionnaire. A healthy eating index (HEI) was developed from Danish food-based dietary guidelines, and was based on eight components: fruits and vegetables, fish, dietary fibre, red meat, saturated fatty acids, sodium, sugar-sweetened beverages, and added sugar. HEI was divided into quartiles with those in the highest quartile having the more optimal diet. A log-linear binomial model was used to estimate the relative risk of the offspring being in the highest quartile of HEI at 14 years if the mother was also ranked in the highest quartile of HEI during pregnancy. Ethics approval was given by the National Committee on Health Research Ethics (H-4-2011-045). Written informed consent was obtained from the children's parents. FindingsAt 14 years old, a child had an increased probability of being in the highest quartile of HEI if the mother was ranked in the highest quartile of HEI during pregnancy (relative risk 2·05, 95% CI 1·91–2·20). Adjustment for maternal pre-pregnancy body-mass index (BMI), parity, level of education, smoking, and energy intake of offspring did not affect the estimates, nor did analyses within strata of pre-pregnancy BMI, level of education, and maternal smoking. InterpretationThis is the first study, to our knowledge, to examine associations between maternal dietary habits during pregnancy and offspring dietary habits during adolescent years in a large contemporary cohort. Diet quality during pregnancy was associated with the quality of offspring diet at age 14 years. These findings could be driven by biological factors, social determinants, or both. Nevertheless, the findings have major potential implications for the developmental origins of health and disease since they emphasise the importance of taking both maternal and offspring diet into account when studying dietary causes of adult diseases. FundingThe study was supported by grants from the Danish Research Council, number 09-067124 (Centre for Fetal Programming) and number 09-075611; and by Nordea-fonden, number 02-2013-2014. The funders had no role in the design of the study, collection, analysis, interpretation of data, or writing of the abstract.

Risk of Birth Abnormalities in the Offspring of Men With a History of Cancer: A Cohort Study Using Danish and Swedish National Registries

BackgroundThe potential mutagenic effects of cancer therapies and the growing number of young male cancer survivors have given rise to concern about the health of their offspring.MethodsWe identified all singleton children born alive in Denmark between 1994 and 2004 and in Sweden between 1994 and 2005 (n = 1 777 765). Of the 8670 children with a paternal history of cancer, 8162 were conceived naturally and 508 were conceived using assisted reproductive technologies (ARTs) (in vitro fertilization or intracytoplasmatic sperm injection). Of the 1 769 0795 children without a paternal history of cancer, 25 926 were conceived using ARTs. Associations between paternal history of cancer and risk of adverse birth outcomes of children conceived naturally or by ARTs were investigated using log-linear binomial models, yielding risk ratios (RRs) with 95% confidence intervals (CIs). All statistical tests were two-sided.ResultsThe offspring of male cancer survivors were more likely to have major congenital abnormalities than the offspring of fathers with no history of cancer (RR = 1.17, 95% CI = 1.05 to 1.31, P = .0043, 3.7% vs 3.2%). However, the mode of conception (natural conception or ARTs) did not modify the association between paternal history of cancer and risk of congenital abnormalities (natural conception, RR = 1.17, 95% CI = 1.04 to 1.31; ARTs, RR = 1.22, 95% CI = 0.80 to 1.87, Pinteraction = .84).ConclusionWe observed a statistically significant but modest increase in the risk of major congenital abnormalities among offspring of males with a history of cancer, independent of the mode of conception.

Different Methods to Calculate Effect Estimates in Cross-sectional Studies

According to results from the epidemiological literature, it can be expected that the prevalence odds ratio (POR) and the prevalence ratio (PR) differ with increasing disease prevalence. We illustrate different concepts to calculate these effect measures in cross-sectional studies and discuss their advantages and weaknesses, using actual data from the ISAAC Phase III cross-sectional survey in Münster, Germany. We analyzed data on the association between self-reported traffic density and wheeze and asthma by means of the POR, obtained from a logistic regression, and the PR, which was estimated from a log-linear binomial model and from different variants of a Poisson regression. The analysis based on the less frequent disease, i.e. asthma with an overall prevalence of 7.8%, yielded similar results for all estimates. When wheezing with a prevalence of 17.5% was analyzed, the POR produced the highest estimates with the widest confidence intervals. While the point estimates were similar in the log-binomial model and Poisson regression, the latter showed wider confidence intervals. When we calculated the Poisson regression with robust variances, confidence intervals narrowed. Since cross-sectional studies often deal with frequent diseases, we encourage analyzing cross-sectional data based on log-linear binomial models, which is the 'natural method' for estimating prevalence ratios. If algorithms fail to converge, a useful alternative is to define appropriate starting values or, if models still do not converge, to calculate a Poisson regression with robust estimates to control for overestimation of errors in the binomial data.