Table 3.
Dominance analysis results: explanatory power of each factor to screening positive for depression among women of reproductive age (n 2599) participating in the Food and Agricultural Approaches to Reducing Malnutrition (FAARM) trial in rural north-eastern Bangladesh, 2015
| Contributing variable | Non-peripartum women (n 1440)* Overall fit statistic (pseudo R2) = 0·2654 |
Peripartum women (n 1040)* Overall fit statistic (pseudo R2) = 0·2898 |
||
|---|---|---|---|---|
| Rank | Relative contribution to overall fit | Rank | Relative contribution to overall fit | |
| Cluster | 1 | 0·0747 | 1 | 0·0797 |
| Interviewer | 2 | 0·0708 | 2 | 0·0748 |
| Household food insecurity (HFIAS) | 3 | 0·0480 | 3 | 0·0599 |
| Household food consumption (FCS) | 4 | 0·0165 | 5 | 0·0110 |
| Social support | 5 | 0·0087 | 4 | 0·0146 |
| Low dietary diversity (<5/13 groups) | 6 | 0·0070 | 6 | 0·0099 |
| Age | 7 | 0·0070 | 13 | 0·0042 |
| Mobility | 8 | 0·0066 | 11 | 0·0046 |
| Age at first marriage | 9 | 0·0065 | 8 | 0·0063 |
| Wealth quintile | 10 | 0·0043 | 10 | 0·0048 |
| CED (BMI < 18·5 kg/m2) | 11 | 0·0031 | not used for peripartum women | |
| Live births | 12 | 0·0031 | 12 | 0·0045 |
| Interpersonal communication | 13 | 0·0023 | 15 | 0·0007 |
| Religion | 14 | 0·0023 | 7 | 0·0065 |
| Education | 15 | 0·0021 | 9 | 0·0050 |
| Household size | 16 | 0·0015 | 14 | 0·0027 |
| Decision-making capacity | 17 | 0·0009 | 16 | 0·0006 |
| Anaemia | 18 | 0·0001 | not used for peripartum women | |
HFIAS, Household Food Insecurity and Access Scale; FCS, Food Consumption Score; CED, chronic energy deficiency.
*
Missing values. NPW: ninety-three observations dropped due to missing values for anaemia; two observations dropped due to missing dietary diversity values; nine observations dropped due to missing BMI. PW: three observations dropped due to missing time since birth information.