Table 1.
Summary of papers included: child nutrition
| Authors (year) | Location | Study design | Age range | n | P‐ or S‐BI | BI/IPI categories | Outcomes assessed | Control for confounders | Findings | Quality score (1–5) |
|---|---|---|---|---|---|---|---|---|---|---|
| Aerts et al. (2004) | Brazil | CS | <5 years | 3 355 | P‐BI S‐BI | <24; ≥24 months | Stunting | Maternal age, education, parity, socio‐economic status, employment, child age, sex, gestational age, birthweight, child care | Stunting more likely with P‐BI <24 months (AOR = 1.69; CI 1.13–2.53) and with S‐BI <24 months (AOR = 1.91; CI 1.16–3.13) | 3 |
| Afifi (1985) | Egypt | P‐Coh | <1 year | 162 | P‐BI | <2; 2–3; >3 years | Four growth ‘factors’: growth velocity, size factor, length, soft/ bony tissue | Maternal age, education, social sanitary index, maternal height; infant sex, age of solid‐food introduction. (NR: child age) | Previous BI positively associated with overall size factor, but not with growth velocity, soft/bony tissue or infant length factors | 3 |
| Bertrand et al. (1988) | Zaire | CS | <5 years | 2 038 | P‐BI | <24; >24 months | HAZ, WAZ, WHZ, wasting, stunting, underweight | Maternal age, parity, family income, housing type and quality, BF after BI (current yes/no), child sex and age | No association between BI and HAZ or WHZ; association with WAZ was marginally significant | 4 |
| Boerma & Bicego (1992) | 17 DHS: Bolivia, Brazil, Burundi, Colombia, Dom. Rep., Egypt, Ghana, Guatemala, Mali, Morocco, Senegal, Sri Lanka, Thailand, Togo, Tunisia, Uganda, Zimbabwe | CS | <24 months | Varies by site | P‐BI | <24; 24–35; ≥36 months | Stunting, underweight | Maternal age and education, parity, household assets, floor material; infant sex, child age | Stunting: BI < 24 months at higher risk in 1/17 countries (Zimbabwe, RR 1.47) and lower in 1 country (Dom. Rep, RR 0.72); BI ≥ 36 months at lower risk in 5/17 countries (Morocco, RR 0.70; Tunisia, RR 0.56; Sri Lanka RR 0.68; Colombia, RR 0.50; Dom. Rep., RR 0.46). Underweight: BI < 24 months at higher risk in 2/17 countries (Ghana, RR 1.54; Thailand, RR 1.60). BI ≥ 36 months at lower risk in 1/17 countries (Dom. Republic, RR 0.38). (confidence intervals not reported) | 3 |
| Bøhler et al. (1995) | Bhutan | P‐Coh and CC (subsequent pregnancy within 18–30 months vs. not) | <36 months | 86 | S‐IPI | 18–30; >30 months | Growth velocity during first 6 months of subsequent pregnancy (P‐Coh); WAZ, MUAC, weight gain (CC) | Maternal age, education, parity, father’s occupation, arable land owned; CC subjects matched for child age. (NR: child sex) | No association between subsequent IPI and growth velocity in cohort study. In case–control study, lower weight gain & greater risk of low MUAC found in children whose mothers became pregnant. Weight gain difference most apparent during 1st trimester of pregnancy | 3 |
| Bøhler (1996) | Bhutan (follow‐up to 1995 study) | P‐Coh and CC | ∼4.5 years | 76 (P‐Coh); 66 (CC) | S‐BI | 18–30; >30 months; also as continuous variable | WAZ, HAZ, WHZ, MUAC | Maternal age, parity, father’s occupation, arable land owned, previous BI. (NR: child age) | No association between subsequent BI and any of the anthropometric outcomes. Case–control study yielded same results as regression analysis | 4 |
| Clark (1992) | Guatemala | CS | <1 year | 301 | P‐IPI | ≤12; >12 months | Weight change (0–12, 0–6, 6–12 months) | Maternal age, parity, income, height, BF duration after BI, infant sex, initial weight. (NR: child age) | Short previous IPI associated with decreased growth from 0–12 and 0–6 months, but no association with growth from 6–12 months | 5 |
| Fletcher et al. (1992) | Jamaica | CC | 6–48 months | 612 | P‐BI | <24; ≥24 months | Underweight (<80% median weight for age) | Prenatal care, household possessions, birthweight. (NR: maternal age, education, parity, child sex and age) | BI < 24 months associated with greater risk of underweight (AOR = 1.8; CI 1.02–3.21) | 3 |
| Goldberg (1985) | Senegal | CS | 2–4 years | 979 | P‐BI S‐BI | ≤24; >24 months | WHZ, HAZ | Prenatal care, maternal age, parity, dwelling type, husband’s occupation, ethnicity, child sex | Short (≤24 months) previous BI associated with greater risk of low HAZ (AOR = 2.4), but not associated with low WHZ. Short subsequent BI associated with lower risk of low HAZ (AOR = 0.26), but not associated with low WHZ. (confidence intervals not reported) | 2 |
| Huttly et al. (1992) | Brazil | CS | ∼19 months | 3 587 | P‐BI | <18; 18–23; 24–35; 36–47; 48–71; >71 months | WHZ and HAZ | Maternal age, education, parity, income, race, child age. (NR: child sex, gestational age) | Previous BI positively associated with WHZ and HAZ. Same trend when limiting analysis to those with gestational age data | 3 |
| Madzingira (1995) | Zimbabwe – DHS | CS | 3–60 months | No information | P‐BI | Continuous | WAZ, HAZ, WHZ | Maternal education, parity, income, sanitation, BF duration after BI, child age, birthweight | Previous BI not associated with any of the anthropometric outcomes | 4 |
| Mozumder et al. (2000) | Bangladesh | CS | 6–39 months | 1 887 | P‐BI & S‐BI | P: none, ≤18; 19–24; 25–36; 37–48, ≥49 months S: none; ≤18; 19–24; 25–36; ≥37 months | Low weight for age (<60% and <70% of NCHS median) | Maternal age, education, parity, housing unit area, child age, sex, diarrhoea | Short subsequent BI (≤24 vs. >25 months) associated with greater risk of low weight for age (OR 1.6; CI 1.09–2.23). Previous BI not associated with low weight for age | 3 |
| Rajaram et al. (2003) | India | CS | <5 years | 7 473 | P‐BI | <24; 24–47; ≥48 months | Stunting, wasting, underweight | Maternal age, education, parity, rural/urban, source of lighting, child age, sex | No association between BI and stunting or wasting. In one region, underweight less likely with BI ≥ 48 months (OR = 0.46; CI 0.29–0.74) | 3 |
| Ricci & Becker (1996) | Philippines | CS | <30 months | 14 189 | P‐BI | <24; ≥24 months | Stunting, wasting | Prenatal care, household possessions, house materials, maternal age, education, BF after BI (current yes/no), child age, sex, birthweight | BI < 24 months associated with higher risk of stunting in 3 age intervals (0–5, 6–11, 12–29 months) and both urban and rural locations (AOR 1.13–1.67). BI not associated with wasting. (confidence intervals not reported) | 3 |
| Roy (2000) | Bangladesh | P‐Coh | <5 years | 8 881 | P‐BI S‐BI | None; <24; ≥24 months | Low MUAC (<−3 SD) | Maternal education, family dwelling space, religion, BF duration after BI, child sex, diarrhoea | Previous BI: no association with MUAC. Subsequent BI: lower risk of low MUAC with BI ≥ 24 months (AOR 0.83; CI 0.70–0.99) | 2 |
| Russell (1976) | Guatemala | P‐Coh | <5 years | 643 | P‐BI S‐BI | 12–17; 18–23; 24–29; 30–35; 36–41; ≥42 months | Height at 3 years, interpolated from growth curves | Parity, parental height, child sex. (NR: child age) | Subsequent BI positively associated with height. Previous BI positively associated with height only in males | 3 |
| Rutstein (2005) | 14 DHS: Bangladesh, Bolivia, Cote d’Ivoire, Egypt, Ghana, Guatemala, India, Indonesia, Kenya, Morocco, Nepal, Nigeria, Peru, Philippines, Tanzania, Uganda, Zambia | CS | <5 years or <4 years or <3 years (depending on country) | 94 376 | P‐BI | <18; 18–23; 24–29; 30–35; 36–41; 42–47; 48–53; 54–59; ≥60 months | Stunting, wasting, underweight | Maternal age, education, urban/rural, socio‐economic status, parity, prenatal care, wantedness of pregnancy, child age, sex, breastfeeding | Stunting less likely with longer BI in 6 countries (Bangladesh, Guatemala, India, Kenya, Peru, Zambia) and underweight less likely with longer BI in 6 countries (Bolivia, Guatemala, India, Kenya, Nigeria, Peru). No association of BI with wasting | 4 |
| Sekiyama & Ohtsuka (2005) | Indonesia | CS | 5–12 years | 310 | P‐BI | Continuous | HAZ, WAZ | Maternal age, education, parity, height, weight, socio‐economic status, healthcare utilization, child age, sex, birthweight | BI positively associated with HAZ and WAZ | 4 |
| Som et al. (2006) | India | CS | <3 years | 1 789 | P‐BI | <24; 24–47; ≥48 months | Stunting, wasting, underweight | Maternal age, education, parity, socio‐economic status, ethnicity, urban/rural, child age, sex | No consistent association between BI and stunting, wasting or underweight | 2 |
| Sommerfelt (1991) | 19 DHS (17 also reported by Boerma & Bicego) | CS | 3–36 months | Varies by site | P‐BI | None; <24; 24–35; 36–47; >47 months | HAZ | Maternal education, parity, household assets, sanitation, child sex, child age | No association between HAZ and previous BI in Nigeria or Trinidad and Tobago* | 3 |
| Thaver et al. (1990) | Pakistan | CS | <1 year | 211 | P‐BI | <24; >24 months | Underweight (not defined) | Prenatal care, maternal age, parity, housing type and materials, family possessions, mortality of previous sibling, BF duration after BI, child age | Short BI not associated with underweight | 4 |
| Thuita et al. (2005) | Kenya | CS | 6–36 months | 347 | P‐BI | Continuous | HAZ, WAZ, WHZ | Maternal age, education, parity, socio‐economic status, employment, BMI, prenatal care, child sex | BI positively associated with WAZ but not HAZ or WHZ | 2 |
AOR, adjusted odds ratio; BF, breastfed; BI, birth interval (P‐BI, previous birth interval; S‐BI, subsequent birth interval); CC, case–control; CI, 95% confidence interval; CS, cross‐sectional; DHS, Demographic and Health Surveys; Dom. Rep., Dominican Republic; HAZ, height‐for‐age z‐score; IPI, interpregnancy interval; MUAC, mid‐upper arm circumference; NCHS, National Center for Health Statistics; NR, not relevant (not a confounder); OR, odds ratio; P‐Coh, prospective cohort; RR, relative risk; Stunting, HAZ < −2; Underweight, WAZ < −2; Wasting, WHZ < −2; WAZ, weight‐for‐age z‐score; WHZ, weight‐for‐height z‐score; *Results from the 17 other countries were reported in the Boerma & Bicego paper.