The majority of groceries purchased by US households are industrially processed, yet it is unclear how processing level influences diet quality. We sought to determine if processing level is associated with diet quality of grocery purchases.

We analysed grocery purchasing data from the National Household Food Acquisition and Purchase Survey 2012–2013. Household grocery purchases were categorized by the NOVA framework as minimally processed, processed culinary ingredients, processed foods or ultra-processed foods. The energy share of each processing level (percentage of energy; %E) and Healthy Eating Index-2015 (HEI-2015) component and total scores were calculated for each household’s purchases. The association between %E from processed foods and ultra-processed foods, respectively, and HEI-2015 total score was determined by multivariable linear regression. Foods purchased by households with the highest v. lowest ultra-processed food purchases and HEI-2015 total score <40 v. ≥60 were compared using linear regression.

USA.

Nationally representative sample of 3961 households.

Processed foods and ultra-processed foods provided 9·2 (se 0·3) % and 55·8 (se 0·6) % of purchased energy, respectively. Mean HEI-2015 score was 54·7 (se 0·4). Substituting 10 %E from minimally processed foods and processed culinary ingredients for ultra-processed foods decreased total HEI-2015 score by 1·8 points (β = −1·8; 95 % CI −2·0, −1·5). Processed food purchases were not associated with diet quality. Among households with high ultra-processed food purchases, those with HEI-2015 score <40 purchased less minimally processed plant-foods than households with HEI-2015 score ≥60.

Increasing purchases of minimally processed foods, decreasing purchases of ultra-processed foods and selecting healthier foods at each processing level may improve diet quality.

To understand the effects of interviewers on the responses they collect for measures of food security, income and selected survey quality measures (i.e. discrepancy between reported Supplemental Nutrition Assistance Program (SNAP) status and administrative data, length of time between initial and final interview, and missing income data) in the US Department of Agriculture’s National Household Food Acquisition and Purchase Survey (FoodAPS).

Using data from FoodAPS, multilevel models with random interviewer effects were fitted to estimate the variance in each outcome measure arising from effects of the interviewers. Covariates describing each household’s socio-economic status, demographics and experience in taking the survey, and interviewer-level experience were included as fixed effects. The variance components in the outcomes due to interviewers were estimated. Outlier interviewers were profiled.

Non-institutionalized households in the continental USA (April 2012–January 2013).

Individuals (n 14 317) in 4826 households who responded to FoodAPS.

There was a substantial amount of variability in the distributions of the outcomes examined (i.e. time between initial and final interview, reported values for food security, individual income, missing income) among the FoodAPS interviewers, even after accounting for the fixed effects of the household- and interviewer-level covariates and removing extreme outlier interviewers.

Interviewers may introduce error in food acquisition survey data when they are asked to interact with the respondents. Managers of future surveys with similarly complex data collection procedures could consider using multilevel models to adaptively identify and retrain interviewers who have extreme effects on data collection outcomes.