Evaluation and interpretation of latent class modelling strategies to characterise dietary trajectories across early life: a longitudinal study from the Southampton Women's Survey

Adiposity trajectories and cardiovascular disease risk in women: a population-based cohort study with a focus on menopausal status

Objectives

A single measurement of adiposity indices could predict the incidence of cardiovascular disease (CVD); nonetheless their long-term pattern and its association with incident CVD are rarely studied. This study aimed to determine distinct trajectories of adiposity indices among participants of Tehran Lipid and Glucose Study (TLGS) and their association with incident CVD. Furthermore, this study aimed to investigate whether this association differed among individuals according to their menopausal status.

Method

A total of 6840 women participated in TLGS, aged 20 years and older were included in this study; they were followed for a median of 16 years. Body mass index (BMI), waist circumference (WC), conicity index (CI) and body roundness index (BRI) were included in the analysis as adiposity indices. The cohort outcome panel of medical specialists identified the CVD outcomes. Trajectory analyses were used to identify homogeneous distinct clusters of adiposity indices trajectories. The association between the trajectory group membership and incident CVD were explored by Cox proportional hazard models, with unadjusted and adjusted model for baseline age, physical activity, smoking status, menopause and family history of CVD.

Results

Three BMI trajectory groups of low, medium, and high and two trajectories for WC, BRI and CI were identified. Adjusted cox proportional hazard models revealed significant associations between the hazard of CVD experience and the high trajectory group of the BMI (HR: 2.06, 95% CI: 1.38-3.07), WC (HR: 2.71, 95% CI: 1.98-3.70), CI (HR: 1.87, 95% CI: 1.26-2.77) and BRI (HR: 1.55-95% CI: 1.12-2.15), compared to the low trajectory group. Subgroup analysis based on the menopausal status of participants showed that the HR of CVD incidences for all of trajectories adiposity indices, except BMI, was statistically significant. Adjusted cox proportional hazard models, in those women not reached menopause during study, revealed that the HR (95% CI) of CVD incidences for high trajectory of BMI, WC, CI and BRI were 2.80 (1.86-7.05); 2.09 (1.40-6.16); 1.72 (1.42-5.61), and 3.09 (1.06-9.01), respectively. These values for those were menopause at the initiation of the study were 1.40 (1.11, 2.53); 1.65 (1.04-2.75); 1.69 (1.01-2.87), and 1.61 (0.98-2.65), respectively.

Conclusion

Our findings suggest that adiposity trajectories, particularly central adiposity index of CI, could precisely predict the CVD risk. Consequently, preventive strategies should be tailored accordingly.

Birth outcomes and early growth patterns associated with age at adiposity rebound: the Ma'anshan birth cohort (MABC) study

OBJECTIVE

Early onset of adiposity rebound (AR) is considered an early indicator of obesity risk. Our objective was to investigate the association of birth outcomes and early physical growth patterns with early AR in children.

METHODS

Study subjects (n = 2705) were enrolled from the Ma'anshan birth cohort (MABC). The body mass index (BMI), head circumference, waist circumference, and body fat were collected. Rapid weight gain (RWG) was defined by the change in weight standard-deviation score in the first two years of life. Group-based trajectory modeling (GBTM) was used to determine children's physical growth trajectories. The age of AR was fitted using fractional polynomial function models.

RESULTS

Children with very high BMI trajectories (RR = 2.83; 95% CI 2.33 to 1.40), rising BMI trajectories (RR = 3.15; 95% CI 2.66 to 3.72), high waist circumference trajectories (RR = 4.17; 95% CI 3.43 to 5.06), and high body fat trajectories (RR = 3.01; 95% CI 2.62 to 3.46) before 72 months of age were at a greater risk of experiencing early AR. Low birth weight (LBW) (RR = 1.86; 95% CI 1.28 to 2.51), preterm birth (PTB) (RR = 1.50; 95% CI 1.17 to 1.93), and small for gestational age (SGA) (RR = 1.37; 95% CI 1.14 to 1.64) associated with increased risk of early AR. Moreover, infants experiencing RWG (RR = 1.59; 95% CI 1.40 to 1.83), low BMI trajectories (RR = 1.27; 95% CI 1.06 to 1.53) and rising BMI trajectories (RR = 1.50; 95% CI 1.22 to 1.84) in the first two years were at higher risk of developing early AR subsequently. Compared to the group with non-early AR, the BMI of children with early AR tended to be lower first (from birth to 6 months of age) and then higher (from 18 to 72 months of age).

CONCLUSIONS

Children with overall high BMI, high waist circumference, and high body fat before 72 months of age are more likely to experience early AR, but infants with low BMI trajectories, rising BMI trajectories and infants experiencing RWG in the first two years of life similarly increase the risk of early AR. These results can help to understand the early factors and processes that lead to metabolic risks.

Diet Quality Trajectories over Adulthood in a Biracial Urban Sample from the Healthy Aging in Neighborhoods of Diversity across the Life Span Study

Limited investigation has been done on diet quality trajectories over adulthood. The main study objectives were to determine the diet quality group trajectories (GTs) over time and to detect changes in a socio-economically and racially diverse middle-aged cohort. Data from three waves of the Healthy Aging in Neighborhoods of Diversity across the Life Span (HANDLS) study were used to determine diet quality with group-based trajectory modeling (GBTM). Three quality indices-the Healthy Eating Index (HEI), the Dietary Inflammatory Index (DII), and the Mean Adequacy Ratio (MAR)-were explored. The rate of change in quality over time was determined by mixed-effects regression analysis. Three diet quality GTs, low, middle, and high quality, were identified for each index and confirmed with spaghetti plots. Within each GT, only small changes in diet quality scores were observed, with improvements for the HEI and DII indices and a slight decline in MAR scores. Weighted kappa values revealed that the DII had better agreement with the HEI-2010 and MAR indices compared with the agreement between the HEI-2010 and MAR. Bayesian estimates revealed that the annualized rate of change in diet quality per person across the GTs was similar. There was minimal change in diet quality over time, regardless of the diet quality index used.