Skip to main content

Table 2 Factors associated with low birth weight in Nepal: Adjusted odds ratio

From: Role of antenatal care and iron supplementation during pregnancy in preventing low birth weight in Nepal: comparison of national surveys 2006 and 2011

Factors

Model 1

Model 2 (NDHS 2006)

Model 3 (NDHS 2011)

Antenatal care

4 or more ANC visit

1.00

1.00

1.00

No ANC visit

2.301 (1.526-3.471)*

3.118 (1.628-5.696)*

1.578 (1.062-2.345)*

1-3 ANC visit

1.513 (0.948-2.412)

2.031 (1.050-3.929)

1.080 (0.700-1.665)

Iron consumption during pregnancy

Yes

1.00

1.00

1.00

No/Do not know

1.839 (1.282-2.636)*

1.840 (1.087-3.116)*

1.490 (1.013-2.193)*

Maternal education

Secondary

1.00

1.00

1.00

No education

0.802 (0.588-1.154)

1.027 (0.615-1.714)

0.717 (0.449-1.147)

Primary

1.491(1.024-2.171) *

2.113 (1.224-3.650)*

1.229 (0.751-2.013)

Higher

0.910 (0.525-1.578)

0.358 (0.127-1.008)

1.130 (0.611-2.092)

Development region

Central

1.00

1.00

1.00

Eastern

1.982 (1.261-3.115)*

1.433 (0.794-2.584)

1.872 (1.138-3.080)*

Western

1.245 (0.683-2.272)

1.244 (0.644-2.404)

1.110 (0.581-2.120)

Mid -western

1.029 (0.501-2.114)

0.709 (0.161-3.122)

1.501(0.905-2.490)

Far-western

1.910 (1.035-3.528)*

2.433 (0.899-6.659)

1.736 (1.059-2.847)*

  1. *:statistically significant. NDHS: Nepal Demographic and Health Surveys.
  2. Model 1: All independent variables of interest shown in Figure 1 (Conceptual framework) were entered in the initial model for pooled data, and 2006 and 2011 survey data. # ANC was replaced by iron (in each models, and the odds ratios were presented in the above mentioned table) as iron supplementation and ANC visits were highly correlated (r = 0.8). Model 2 and Model 3: Effect modification of the survey period were examined including the four interaction terms ANC*Year (F = 2.697; p = 0.069), Iron*Year (F = 0.908; p =0.341 ), Region*Year (F = 0.505; p = 0.732), Education of mothers* Years (F = 2.936, p value = 0.033) separately. As one of the interactions was found statistically significant, we built additional regression models for each year separately; Model 2: Survey 2006 and Model 3: Survey 2011.