Study design, area and population

An analytical cross-sectional study design was used in the present study. The study was conducted in Kalpohini, Kpalsi, Sagnarigu and Wurshe communities in the Sagnarigu municipality of Northern Ghana from August to October, 2022. Eligible participants were mother-child pairs with the children aged 6–23 months who consented to participate in the study. Participants who were not sound minded were excluded from the study.

Sampling

Multi-stage sampling method was used in this study. Four study communities were selected through simple random sampling. All the 79 communities under Sagnarigu Municipality were listed and assigned specific numbers written on pieces of paper, put in a basin, mixed, and picked one at a time with replacement till the number of communities were reached. Systematic random sampling was used to select the households, where the households were selected according to a random starting point but with a fixed interval. This interval was calculated by dividing the total number of households by the desired sample size. In households with two or more eligible children, only one child was selected through simple random sampling. The population proportion to size method was utilized to calculate the sample size for each of the communities.

Sample size

The sample size for this study was calculated using the Cochran’s formula:

N\( =\frac^p(1-p)}^}\) [23]

Where:

N is the sample size.

z is confidence interval (95%) which gives a critical value of 1.96.

p is the estimated proportion of an attribute present in the population. Prevalence of minimum dietary diversity is 35.3% in northern Ghana [16].

M.E is the desired level of precision (5%=0.05).

Hence, N = 351. A 5% contingency was considered to cover up incomplete questionnaires. Hence, N = 369.

Study variablesDependent variable

The dependent variable was MDD.

Independent variables

The independent variables were caregiver characteristics (age, marital status, ethnicity, religion, educational level, occupation, income level, birth interval, antenatal care visits, post-natal care visits, place of delivery, breastfeeding status, timely initiation of breastfeeding and nutritional knowledge) and child characteristics (age, sex and history of illness in the past 2 weeks).

Data collectionPretesting of questionnaire

The questionnaire pretesting was done by administering the questionnaires to 37 (10% of total sample size) caregivers with children aged 6–23 months in the study district; these caregivers did not participate in the study. This allowed for the researchers to fine-tune the questions for clarity and comprehension by the participants.

Socio-demographic characteristics

The pretested semi-structured questionnaire was used to take information on socio-demographic characteristics of the mother including age, marital status, ethnicity, religion, educational level, occupation, income level and sex of child. Data collection was facilitated by two trained research assistants.

Health service utilization and obstetric characteristics

Participants’ characteristics including antenatal care (ANC) visits, post-natal care (PNC) visits, place of delivery and birth interval were also documented.

Caregivers knowledge on dietary diversity and feeding practices

Caregivers’ knowledge on dietary diversity and feeding practices was assessed using a questionnaire adapted from Solomon et al. [24]. The questionnaire consisted of ten knowledge questions. Each correct answer (yes) was assigned a score of 1, while any wrong answer was assigned a zero (0) score. Mothers who got a score of 7 and above out of the ten knowledge questions were deemed to have good knowledge while mothers who got a score of less than 7 were deemed to have poor knowledge [24].

Dietary diversity

Data on MDD was collected using the WHO indicators for assessing IYCF practices [25]. A 24-hour dietary recall, repeated in 20% of random sub-sample [26], was used to obtain food intake information. The 24-hour dietary recall was conducted on two non-consecutive days, one weekday and one weekend [27,28,29]. The respondents were asked to recall all foods eaten and beverages taken by the children in the previous 24 h prior to the interview. The dietary diversity score (DDS) was assessed by assigning a score of 1 to a food group if the child ate any food item from the food group and a score of 0 if no food item from the food group was consumed. A total of 8 food groups [25] were considered in this study: Breast milk; grains, roots and tubers; legumes, nuts and seeds; dairy products; flesh foods, eggs; vitamin A-rich fruits and vegetables and other fruits and vegetables. Consequently, the minimum possible DDS score was 0 (no food group consumed) and the maximum possible DDS score was 8 (all food groups consumed). A child was classified as having achieved the MDD if he/she consumed at least 5 food groups out of the 8.

Anthropometry

Length (cm) of child was measured without footwear using an infantometer (Seca, Germany) and weight (kg) was measured without clothing using a digital weighing scale (Seca, Germany). The child’s age, sex, and measurements of weight and length were used to calculate the following growth indicators: length-for-age (stunting), weight-for-age (underweight) and weight-for-length (wasting) in accordance with the WHO 2006 child growth reference [30]. The cut off point for stunting, wasting and underweight was − 2SD from the median of WHO child growth standard.

Data analysis

The Statistical Package for Social Science (SPSS version 22) was used to analyze the data. The results were reported with descriptive statistics including frequencies, percentages, means, and standard deviations. Chi-square/Fisher exact test was used for bivariate analysis. The factors with p < 0.25 [31] in the bivariate analysis were selected for multivariate logistic regression analysis to determine predictors of MDD. P < 0.05 (at two-tailed test) was considered significant.