We developed an agent-based model using a trial emulation approach to quantify effect measure modification of spillover effects of pre-exposure prophylaxis (PrEP) for HIV among men who have sex with men (MSM) in the Atlanta-Sandy Springs-Roswell metropolitan area, Georgia. PrEP may impact not only the individual prescribed, but also their partners and beyond, known as spillover. We simulated a two-stage randomised trial with eligible components (≥3 agents with ≥1 HIV+ agent) first randomised to intervention or control (no PrEP). Within intervention components, agents were randomised to PrEP with coverage of 70%, providing insight into a high PrEP coverage strategy. We evaluated effect modification by component-level characteristics and estimated spillover effects on HIV incidence using an extension of randomisation-based estimators. We observed an attenuation of the spillover effect when agents were in components with a higher prevalence of either drug use or bridging potential (if an agent acts as a mediator between ≥2 connected groups of agents). The estimated spillover effects were larger in magnitude among components with either higher HIV prevalence or greater density (number of existing partnerships compared to all possible partnerships). Consideration of effect modification is important when evaluating the spillover of PrEP among MSM.

Social scientists use the concept of interactions to study effect dependency. In the causal inference literature, interaction terms may be used in two distinct type of analysis. The first type of analysis focuses on causal interactions, where the analyst is interested in whether two treatments have differing effects when both are administered. The second type of analysis focuses on effect modification, where the analyst investigates whether the effect of a single treatment varies across levels of a baseline covariate. While both forms of interaction analysis are typically conducted using the same type of statistical model, the identification assumptions for these two types of analysis are very different. In this paper, we clarify the difference between these two types of interaction analysis. We demonstrate that this distinction is mostly ignored in the political science literature. We conclude with a review of several applications where we show that the form of the interaction is critical to proper interpretation of empirical results.

Previous studies have suggested that a high intake of sugar-sweetened beverages is positively associated with the risk of a coronary event. However, a few studies have examined the association between sucrose (the most common extrinsic sugar in Sweden) and incident coronary events. The objective of the present study was to examine the associations between sucrose intake and coronary event risk and to determine whether these associations are specific to certain subgroups of the population (i.e. according to physical activity, obesity status, educational level, alcohol consumption, smoking habits, intake of fat and intake of fruits and vegetables). We performed a prospective analysis on 26 190 individuals (62 % women) free from diabetes and without a history of CVD from the Swedish population-based Malmö Diet and Cancer cohort. Over an average of 17 years of follow-up (457 131 person-years), 2493 incident cases of coronary events were identified. Sucrose intake was obtained from an interview-based diet history method, including 7-d records of prepared meals and cold beverages and a 168-item diet questionnaire covering other foods. Participants who consumed >15 % of their energy intake (E%) from sucrose showed a 37 (95 % CI 13, 66) % increased risk of a coronary event compared with the lowest sucrose consumers (<5 E%) after adjusting for potential confounders. The association was not modified by the selected lifestyle factors. The results indicated that sucrose consumption higher than 15 E% (5 % of this population) is associated with an increased risk of a coronary event.

Several N-nitroso compounds (NOC) have been shown to be carcinogenic in a variety of laboratory animals, but evidence of their carcinogenicity in humans is lacking. We aimed to examine the association between NOC intake and colorectal cancer (CRC) risk and possible effect modification by vitamins C and E and protein in a large case–control study carried out in Newfoundland and Labrador and Ontario, Canada. A total of 1760 case patients with pathologically confirmed adenocarcinoma and 2481 population controls were asked to complete a self-administered FFQ to evaluate their dietary intakes 1 year before diagnosis (for cases) or interview (for controls). Adjusted OR and 95 % CI were calculated across the quintiles of NOC (measured by N-nitrosodimethylamine (NDMA)) intake and relevant food items using unconditional logistic regression. NDMA intake was found to be associated with a higher risk of CRC (highest v. lowest quintiles: OR 1·42, 95 % CI 1·03, 1·96; P for trend = 0·005), specifically for rectal carcinoma (OR 1·61, 95 % CI 1·11, 2·35; P for trend = 0·01). CRC risk also increased with the consumption of NDMA-containing meats when the highest tertile was compared with the lowest tertile (OR 1·47, 95 % CI 1·03, 2·10; P for trend = 0·20). There was evidence of effect modification between dietary vitamin E and NDMA. Individuals with high NDMA and low vitamin E intakes had a significantly increased risk than those with both low NDMA and low vitamin E intakes (OR 3·01, 95 % CI 1·43, 6·51; P for interaction = 0·017). The present results support the hypothesis that NOC intake may be positively associated with CRC risk in humans. Vitamin E, which inhibits nitrosation, could modify the effect of NDMA on CRC risk.

Delta-5 (D5D) and delta-6 (D6D) desaturases are key enzymes in PUFA metabolism. Several factors (e.g. hyperglycaemia, hypertension, blood lipids, statins and fatty acids in diet and biological tissues) may influence desaturase activity. The goals were to evaluate the associations between variation in genes encoding these desaturases (FADS1 and FADS2) and blood concentrations of n-6 PUFA and estimated D5D and D6D activities (evaluated as product/precursor ratio), and to investigate whether other factors influencing the activity of desaturases modify these associations. A random sample of 2066 participants from the European Prospective Investigation into Cancer and Nutrition-Potsdam study (n 27 548) was utilised in the analyses. Crude and adjusted associations between rs174546 genotypes (reflecting genetic variation in the FADS1FADS2 gene cluster), n-6 PUFA in erythrocytes and estimated desaturase activities were evaluated using multiple linear regression. Potential effect modification was determined by performing stratified analyses and evaluating interaction terms. We found rs174546 genotypes to be related to linoleic (r2 0·060), γ-linolenic (r2 0·041), eicosadienoic (r2 0·034), arachidonic (r2 0·026), docosatetraenoic acids (r2 0·028), estimated D6D activity (r2 0·052) and particularly strongly to dihomo-γ-linolenic acid (DGLA, r2 0·182) and D5D activity (r2 0·231). We did not observe effect modifications with regard to the estimated D5D activity, DGLA and arachidonic acid (AA) for most of the factors evaluated; however, the genetic effect on D5D activity and DGLA may be modified by the dietary n-6:n-3-ratio (P-values for interaction: 0·008 and 0·002), and the genetic effect on DGLA and AA may be modified by lipid-lowering medication (P-values for interaction: 0·0004 and 0·006). In conclusion, genetic variation in the FADS1 FADS2 gene cluster affects n-6 PUFA profiles in erythrocytes reflecting altered D5D activity.

Countries worldwide, including the Netherlands, recommend that women planning pregnancy use a folic acid supplement during the periconception period. Some countries even fortify staple foods with folic acid. These recommendations mainly focus on the prevention of neural tube defects, despite increasing evidence that folic acid may also influence birth weight. We examined whether periconception folic acid supplementation affects fetal growth and the risks of low birth weight, small for gestational age (SGA) and preterm birth, in the Generation R Study in Rotterdam, the Netherlands. Main outcome measures were fetal growth measured in mid- and late pregnancy by ultrasound, birth weight, SGA and preterm birth in relation to periconception folic supplementation (0·4–0·5 mg). Data on 6353 pregnancies were available. Periconception folic acid supplementation was positively associated with fetal growth. Preconception folic acid supplementation was associated with 68 g higher birth weight (95 % CI 37·2, 99·0) and 13 g higher placental weight (95 % CI 1·1, 25·5), compared to no folic acid supplementation. In these analyses parity significantly modified the effect estimates. Start of folic acid supplementation after pregnancy confirmation was associated with a reduced risk of low birth weight (OR 0·61, 95 % CI 0·40, 0·94). Similarly, reduced risks for low birth weight and SGA were observed for women who started supplementation preconceptionally, compared to those who did not use folic acid (OR 0·43, 95 % CI 0·28, 0·69 and OR 0·40, 95 % CI 0·22, 0·72). In conclusion, periconception folic acid supplementation is associated with increased fetal growth resulting in higher placental and birth weight, and decreased risks of low birth weight and SGA.