This research developed a methodological protocol that allows the triangulation of accelerometer data with time use and food intakes in rural settings of low- and middle-income countries (LMICs). Household and individual surveys were administrated to 120 individuals in rural Ghana (Upper West Region), India (Telangana State) and Nepal (Province #3). The data collection spanned from June 2017 to November 2018 and participants were invited to wear a medical-graded accelerometer device (ActiGraph GT3X+) throughout the day for four full non-consecutive weeks across an agricultural season. Raw accelerometer data has been archived at one-second interval. Daily individual questionnaires collected information on time use and food intake. Datasets can be used as stand-alone, yet richer analysis can be done linking the various sources of information. Such set of data opens potential new avenues to accurately measure energy expenditure in rural livelihood, thereby complementing an important part of human health and nutrition research in LMICs.This research aimed at developing methods and protocols for generating reliable human energy expenditure profiles that can be used in a variety of agricultural settings in developing countries. The research was carried out in Ghana, India, and Nepal and it developed a protocol in which respondents’ physical activity data (accelerometry data) can be integrated with time use and food intake data. The data and methodology developed through this research provide insights into (1) the energy expenditure and time allocation dimension of agricultural practices and innovations and (2) the gender-differentiated intra-household labour and time allocation associated with these practices and innovations. Such insights are critical for the design of nutritionally-sensitive agricultural interventions.

The data was collected using household questionnaires, individual questionnaires, and accelerometer sensors in two communities in each country, one adopting predominantly rain-fed agriculture and one with irrigation infrastructure. In each community 10 households were sampled, stratified by the size of the land being cultivated (smallholder, medium, and large-holder). In the randomly selected households, we invited the head of the household and the spouse to participate in the study. To capture different rural activities throughout an agricultural season, we invited each participant to wear an accelerometry device for four non-consecutive weeks across the agricultural season. These corresponded to four key phases: (1) land preparation, (2) sowing and seeding, (3) maintaining (e.g., activities related to weeding and fertilizing), and (4) harvest.