| Author Year of publication Study design | Year of birth of study children Sample population Setting/data source % single mothers | Maternal employment measure (prevalence %) and comparator | Health indicator | Unadjusted OR (95% CI)* (OR>1 indicates poorer health outcome) | Adjusted OR (95% CI) (OR>1 indicates poorer health outcome) | Variables included in adjusted analysis | Quality rating† |
| Anderson232003Cohort | 1975–1993††Mothers and their children (n = 16650) aged 3–11 yearsNational Longitudinal Survey of Youth, USA27.9% | a) Mother worked ⩾35 hours per week (prevalence not stated) vs never workedb) Mother worked <35 hours per week (prevalence not stated) vs never workedc) Employment intensity: average hours worked per week if working since child’s birth (units of 10 hours)d) Employment duration: number of weeks mother employed since child’s birth (units of 52 weeks) | Child overweight: BMI >95th percentile of CDC growth charts | a) 1.43(CI not calculable)¶b) 1.08(CI not calculable)¶c) No unadjusted analysis d) No unadjusted analysis | a) No adjusted analysisb) No adjusted analysisc) Probit estimates: mothers who work 10 hours more per week increase the likelihood that their children will be overweight by 1.2 percentage pointsStratification by maternal education: estimated increase in childhood overweight per extra 10 hours worked is greater for children of more educated than less educated mothersd) Probit estimates: number of weeks of maternal employment not significantly associated with childhood overweight | a) N/Ab) N/Ac) Block 1§Stratification by maternal educaton:Blocks 1, 2 and 3§d) Blocks 1 and 2§ | II AAA |
| Takahashi421999Cohort | 1989Children (obese, n = 427; non-obese, n = 854) aged 3 yearsToyama prefecture, JapanN/R | Mother’s job, full-time (34.6%) vs not full-time | Child obesity: BMI ⩾18 | 1.33 (p<0.05) | No significant association (OR not reported) | Obesity of mother and father. Child’s physical activity/weight at birth/duration of outdoor playtime/snacking regularity/meal times/BMI at birth/snacking frequency/sleeping time/bedtime/wake-up time/kindergarten attendance/seasoning of food. Caretaker status. | II AAA |
| Melgar-Quinonez432004Cross-sectional | 1993–95Families (n = 238) with a child aged 3–5 years, of low income and Mexican American ethnic originCalifornia, USAN/R | Mother employed outside the household [41%] vs not employed outside the household | Child overweight (BMI ⩾85th percentile of CDC growth charts) or obesity (BMI ⩾95th percentile) | No significant association (OR not reported) | Adjusted analysis does not include maternal employment variable | N/A | II AA |
| The following paper comprised cross-sectional and case-control elements | |||||||
| Lamerz282005Cross-sectional | 1995–96Parents of children (n = 1979) aged 5–7 yearsAachen, Germany11% | a) Maternal employment, full-time (8.9%) vs noneb) Maternal employment, part-time (39.4%) vs nonec) Maternal employment, full or part-time (48.3%) vs noned) Maternal employment, full-time vs part-time | Child obesity: BMI ⩾90th percentile for German children | a) 0.75 (0.41 to 1.37)¶b) 0.62 (0.43 to 0.88)¶c) 0.64 (0.46 to 0.89¶d) 1.21 (0.64 to 2.30)¶ | a) 0.92 (0.49 to 1.75)b) 0.82 (CI not calculable)c) No data availabled) 1.12 (0.58 to 2.17) | Education/BMI of mother and father. Father’s employment. Living space per person (m2). Single parent. Gender | III AAA |
| Lamerz282005Case-control | 1995–96Children (obese, n = 146; non-obese, n = 221) aged 5–7 yearsAachen, Germany11% | a) Maternal employment (55.2%) vs noneb) Working more than 4 hours per day (25.4%) vs nonec) Working >0–4 hours per day (28.6%) vs noned) Working more than 4 hours per day vs >0–4 hours per daye) Working at weekends (19.4%) vs not working | Child obesity (cases = BMI ⩾85th percentile for German children; controls = BMI 40th to 60th percentile for German children) | a) 0.60 (0.38 to 0.94)¶b) 0.78 (0.45 to 1.35)¶c) 0.47 (0.27 to 0.81)¶d) 1.66 (0.88 to 3.10)¶e) 0.79 (0.44 to 1.41)¶ | a) No significant association (OR not reported)b) No significant association (OR not reported)c) No significant association (OR not reported)d) No significant association (OR not reported)e) No significant association (OR not reported) | Education/type of occupation of mother and father. Father’s employment/hours of work/hours of work at weekends. Household net income | III AAA |
| Hawkins242007cohort | 2000–02Children (n = 13113) aged 3 yearsMillennium Cohort Study, UK14% | a) Maternal employment, any since child’s birth(58.2%) vs noneb) Hours worked per week, per 10 hoursc) Duration of employment, per 1 year | Child overweight, including obesity, as defined by IOTF cut-offs for BMI | a) 1.07 (0.97 to 1.18)b) 1.06 (1.02 to 1.09)c) 1.03 (0.99 to 1.07) | a) 1.15 (1.02 to 1.29)b) 1.12 (1.06 to 1.18)c) 0.97 (0.91 to 1.04) | Birth weight.Maternal ethnic group/highest academic qualification/age at first live birth/lone motherhood status/pre-pregnancy body size/smoking during pregnancy** | II AAA |
| Kasemsup292006Cross-sectional | <2003††Mothers of overweight (n = 35) or normal weight (n = 45) children aged 3–5 years, of Hmong ethnic originMinneapolis/St Paul metropolitan area, USAN/R | Mother employed part- or full-time (48%) vs not employed | Child overweight: BMI ⩾95th percentile of growth charts | 2.50 (0.98 to 6.37)¶ | No adjusted analysis | N/A | I A |
BMI, body mass index; CDC, Centers for Disease Control and Prevention; CI, confidence intervals; IOTF, International Obesity Task Force; N/A, not applicable; N/R, not recorded; OR, odds ratio.
*95% confidence intervals, unless otherwise stated.
†III = strong internal validity (four or more domains strong or three domains strong and none weak); I = weak internal validity (three or more domains weak); II = moderate internal validity (those not fulfilling criteria for strong or weak). AAA = findings likely to be applicable across a broad range of populations and settings (but may need appropriate adaptation); AA = findings applicable only to populations and settings included in the study – success of broader application is uncertain; A = applicable only to populations and settings included in the study.
‡Estimated year of birth of study children.
¶Odds ratio calculated by review authors using data derived from the published paper.
§Block 1 variables = average hours worked per week if working since child’s birth, number of weeks worked since child’s birth, mother reported height and weight; Block 2 variables = Black non-Hispanic, Hispanic, mother’s education, child was first born, number of children, child’s birthweight, child’s and mother’s age in years, year of survey, education levels of the mother’s parents, whether mother’s parents were present when she was 14, whether the child is female; Block 3 variables = child was breastfed, mother’s BMI, average family income since birth, child’s life mother was married.
**Employment measures b) and c) additionally adjusted for employment [maternal hours worked and duration, partner hours worked and duration].