Table 1.

Summary of key findings from studies of maternal diet during pregnancy [5–10, 13]

Study	Country	Participants	Key findings
Jones et al. 2005 [5]	Australia (Southern Tasmania)	8-year-old male and female children with adequate maternal dietary information during 3rd trimester (n = 173) Estimated mean Ca intake for gravidae was 1905 mg/day (SDa 1310 mg)	+ Assoc between maternal Mg/P density in diet and lumbar spine (LS)/femoral neck (FN) BMD of children + Assoc between K density and LS BMD − Assoc between fat density and LS BMD + Assoc between Mg/P/protein density and total body (TB) BMD − Assoc between fat density and TB BMD (nb all P < 0.05) Linear regression model: no constituent sig assoc with FN BMD or TB BMD. P remained sig + assoc with LS BMD, fat remained − assoc
Yin et al. 2010 [9]	Australia (Southern Tasmania)	16-year-old male and female children with adequate maternal dietary information during 3rd trimester (n = 216) Estimated mean Ca intake for gravidae was 1677 mg/day (SD 841 mg)	+ Assoc between maternal Ca/Mg/P density in diet and LS BMD of children (all P < 0.05) Maternal milk intake + assoc with LS BMD (P < 0.05) − Assoc between fat density and FN BMD (P < 0.05) and LS BMD (P < 0.01) No nutrients assoc with TB BMD in either univariate or multivariate analysis
Petersen et al. 2015 [13]	Denmark	<16-year-old male and female children of gravidae who received semi-quantitative food frequency questionnaires (FFQs) at week 25 of gestation (n = 53,922)	The Western dietary pattern has high intake of fat meat and potatoes and low intake of fruit and vegetables The Western pattern was assoc with forearm fractures in offspring before their 16th birthday High maternal consumption of artificially sweetened drinks increased forearm fractures by 12 % in children
Ganpule et al. 2006 [7]	India	6-year-old male and female children of gravidae who had nutritional status assessed at week 18 and week 28 of gestation (n = 698) Estimated mean Ca intake for gravidae was 274 mg/day at week 18 and 268 mg/day at week 28	Larger parental bone mineral content (BMC) – for both parents – predicted higher TB and/or spine BMC of children (Pearson correlation coefficient P < 0.001). In univariate analyses, maternal frequency of milk intake was + assoc with TB BMD (P < 0.05), as was milk product intake (P < 0.05) In multiple regression analyses, larger parental BMC (for both parents), lower maternal parity and higher maternal intakes of milk and milk products at week 28 of gestation predicted higher TB and/or spine BMC in children
Heppe et al. 2013 [10]	Netherlands	6-year-old male and female children of gravidae who received FFQs at 13.5 weeks of gestation (IQRb 3.3) (n = 2819) Estimated mean Ca intake for gravidae was 1108 mg/day (SD 311 mg)	Higher maternal Ca/P intake assoc with higher BMC and BMD of children (all P-trend < 0.005). No assoc with maternal Mg intake Higher maternal protein intake assoc with higher BMC (P-trend = 0.02) and BMD (P-trend < 0.001) Maternal carbohydrate intake inversely assoc with BMC and BMD (both P-trend = 0.02) No assoc between maternal total energy intake or fat intake and bone mass of children (P-trend > 0.05)
Tobias et al. 2005 [6]	UK	9-year-old male and female children of gravidae who received FFQs at week 32 of gestation (n = 4588) Estimated mean Ca intake for gravidae was 952 mg/day (SD 275 mg)	Maternal Mg intake assoc. with TB BMC (β c = 4.9, 95 % CI 7.4–23.1 g) and BMD (β = 4.9, 2.5–7.3; g/cm2 × 103) (P < 0.001). After adjustment for child height, assoc no longer apparent Maternal P intake assoc with spinal BMC (β = 1.8, 0.8–2.9; g) and BMD (β = 10.5, 4.9–16.0; g/cm2 × 103) (P = 0.001). After adjustment for child weight, assoc is no longer apparent Assoc between maternal dietary folate intake and spinal BMC adjusted for bone area (β = 0.55, 0.16–0.94; g; P = 0.006), still evident after adjustment for height and weight
Cole et al. 2009 [8]	UK	9-year-old male and female children of gravidae who received FFQs at week 15 and week 32 of gestation (n = 198) Estimated mean Ca intake for gravidae was 952 mg/day (SD 275 mg)	The ‘prudent’ diet pattern is characterized by high intakes of fruit, vegetables and wholemeal bread, rice, and pasta, and low intakes of processed foods. Higher prudent diet score at week 32 of gestation assoc with higher intakes of protein and Ca + Assoc between maternal prudent diet score and children’s whole body bone area (r = 0.24, P = 0.001), BMC (r = 0.23, P = 0.001), and areal BMD (r = 0.15, P = 0.02), after adjustment for age and sex

Assoc association, sig significantly

^aStandard deviation

^bInterquartile range

^cStandardised regression coefficients which quantify the absolute change in the dual-energy X-ray absorptiometry (DXA) outcome per standard deviation change in maternal dietary intake for the constituent in question