Table 3.
Associations between meteorology and birth weight.
| Reference | Seasonal or Meteorological Variable | Outcome | Setting (climate type a), study period | Exposure metric | Study Design | Inclusion criteria | Statistical method | Population size | Summarized Main Results | Confounders adjusted for/other comments |
|---|---|---|---|---|---|---|---|---|---|---|
| [81] | Season of birth (month) |
Term birth weight (continuous variable) | Warsaw, Poland (NT), May 2004-April 2005 | - Year divided into four seasons: Spring (April–June); Summer (July–September); Autumn (October–December) and Winter (January-March) |
Hospital based cohort study | All singleton live births after 36 weeks of pregnancy | One way analysis of variance of birth weight transformed to z score Weighted Spearman rank correlation |
N = 10,631 | - Average Z-scores for birth weight associated with month of birth for boys (p = 0.01) respectively, and for girls (p < 0.01). - Peak Z-score values for boys born in October with a trough in March. Peak Z-score values for girls born in July and August with a trough in April. -No association between birth weight and season of birth. |
None |
| [78] | Seasonality of conception (month) Temperature Humidity Rainfall Daylight |
Term birth weight (continuous variable) | Istanbul, Turkey (NT), 1992–2003 | - Women were divided into four groups according to season of last monthly period - Year divided into four seasons: Spring, Summer, Autumn and Winter - Mean daily temperature (°C) and humidity (%), total daily rainfall (mm) and daily duration of daylight (hours) for each trimester of pregnancy |
Hospital based cohort study | All live births after 36 weeks of gestation, except multiple pregnancies | Stepwise multiple linear regression | N = 3,333 | - Women who conceived in winter and spring were exposed to higher temperatures during the second trimester and delivered babies with higher birth weights than those who conceived in summer and autumn. Regression parameter for “Temperature to which the subject was exposed during the second trimester (°C)”: 0.001 multiples of the mean. The mean being about 3,700 g, The gain would be about 3.7 g per °C.) - No association between birth weight and humidity, rainfall, and daylight in any trimester. |
maternal age and parity, mode of delivery, sex |
| [80] | Seasonality of birth (month) Rainfall Sunshine Temperature |
Term birth weight (continuous variable) | Northern, Ireland (NT), 1971–1986 | - Year divided monthly -Mean daily maximum and minimum temperatures, rainfall, and hours of bright sunshine for each pregnancy trimester |
Population based cohort study | Singleton live births after 36 completed wks of gestation | Linear regression | N = 418,817 | - The lowest adjusted mean birth weights were 25.5 g, 29.6 g, and 31.6 g lower in May, June, and July, respectively, than in January - In females, an increase of 1 °C in the mean daily maximum temperature during the second trimester was associated with an increase in mean birth weight of 3.5 g. (SE 0.88) -In males 1.02 (SE 0.88) -No significant association for other trimesters or for rainfall, sunshine, or mean daily minimum temperature |
-Year of birth, duration of gestation, maternal age, number of previous pregnancies, sex, and social class |
| [89] | Seasonality of birth (rainy vs. dry) |
Term birth weight (continuous variable) Term low birth weight (<2,500 g) |
Morogoro, Tanzania (T), | N/A | Hospital based cohort study | All live singleton babies at full term gestation | N/A | N = 19,783, including 2,354 low birth weight infants | - Mean birth weight low during the rainy season and high during the dry season - Low birth weight incidence higher during the rainy than the dry season |
Food intake, energy expenditure |
| [87] | Seasonality of birth (4 seasons) |
Term birth weight (continuous variable) | 12 cities in the USA (NT) 1959–1965 | - Year divided into four seasons: winter (December, January, February); spring (March, April, May); summer (June, July, August); and fall (September, October, November) | Multi- hospital based cohort study | All live births at full term gestation | ANOVA Multiple linear regression |
N = 24,325 | Infants born in fall had lower birth weight than those born in winter (t test = 2.15, p = 0.03) and spring (t test = 2.48, p = 0.01), but no association remained after adjustment for confounders. | Sex, race, maternal age, maternal education, maternal BMI, first born, ever breast fed, weight gain in first four months |
| [79] | Seasonality of birth (month) |
Term birth weight (continuous variable) | Chile (entire country) (NT), 1987–2007 | Year divided monthly and regionally: North, Central-coast, Central-interior and South | Population based cohort study | All live-born singletons with gestations between 37 and 41 weeks in study period | Multivariate regression | N = 4,968,912 | Birth weight has a bimodal peak in spring (p < 0.001) and fall (p < 0.001) and a pronounced nadir in winter and smaller nadir in summer | Maternal age, marital status, college education, urban region |
| [86] | Seasonality of birth (4 seasons) |
Birth weight (continuous variable) | Rome, Italy and Sassary, Italy (NT), January 1993–December 1996 | - Year divided into four seasons: winter, spring, summer, autumn | Hospital based cohort study | All live births | Variance analysis | N = 5,291 | - Birth weight is significantly lower in infants born in winter than in autumn. - mean differenceis 327 g (p < 0.003); after correction for multiple comparison (p < 0.02) |
Population and gestational duration, season of birth (with birth weight as dependent variable) |
| [54] | Seasonality of birth (4 seasons) Temperature |
Birth weight (continuous variable) Low Birth Weight (<2,500 g) |
Greece (entire country) (NT), 1999–2003 | - Year divided into four seasons: winter (December–February), spring (March–May), summer (June–August), and autumn (September–November) - mean air temperature during month of birth |
Population based cohort study | All Greek citizens born or deceased during the period of study | Chi square Tests for contrasts in low birth weight probability between seasons Pearson’s R for association of continuous birth weight with temperature |
N = 516,874 born N = 554,101 died |
- Infants born during autumn and winter had higher birth weight than those born in other seasons of the year - Low birth weight rates were lower (p < 0.05) for infants born during the autumn and winter seasons. -Mean air temperature during the month of birth associated with birth weight r = −0.218 (p < 0.001) |
None |
| [91] | Temperature regime (climate, not seasonal variation) reflected by heat stress (humidity and temperature) | Birth weight (continuous variable) | 140 populations from the WHO (1992) population data | Heat stress index considering yearly average of maximum daily temperature and afternoon humidity | Pooled analysis of population based studies | Population with specific data on birth weight and thermal climate. | Linear regression | 140 populations | - Significant correlation between heat stress and birth weight R2 = −0.59 (p < 0.001) | Data on both birth weight and heat stress were reduced to an annual average value, thus ignored seasonal variation in climate |
| [90] | Temperature regime (climate, not seasonal variation) | Birth weight (continuous variable, log-transformed) | 63 countries from the WHO (1992) population data 1971–2000 | Climate characterized by the mean of daily minimum and maximum temperature from the coolest and warmest months, respectively | Ecological study | Linear regression | 63 countries, number of births not provided | Overall reductions in BW at increasing mean temperatures vary from 0.44% per °C in temperature range 0–5 °C to 1.05% per °C in the temperature range 20–25 °C, subject to adjustment for variation in nutrition, altitude and age of motherhood. | Altitude, prevalence of under-nourishment, obesity, mean age at motherhood, fertility rate, malaria prevalence, geographic origin. | |
| [93] | Extreme temperatures | Birth weight (continuous variable) | USA (entire country), 1972–1988 | Number of days within each pregnancy trimester that fall into different bins of daily average temperature (average of maximum and minimum temperature) <25° F, 25°–45° F, 45°–65° F, 65°–85° F, >85° F). Aggregated at the county level |
Cohort study | Mothers aged 16–45, 48 continental states + DC | Linear regression | N = 37,100,000 | As compared to 45–65° F, each additional day <25° F is associated with a −0.000025 (95% CI: −0.00001; −0.00004) detriment in log birth weight each additional day>85° F is associated with a −0.000025 (95% CI: −0; −0.00005) detriment in log birth weight Linear relationships for temperature exposure during the 2nd and 3rd trimesters: <25° F is associated with a 0.000025 (95% CI: 0.00001; 0.00004) increment in log birth weight each additional day> 85° F is associated with a −0.000075 (95% CI: −0.00006; −0.000012) detriment in log birth weight |
Smooth function for the date of conception. Conditioning by county and year, mother’ s age, fertility history, educational level and marital status Inverse U-shaped dose response relationship between log birth weight and number of days falling within the different bins during the first trimester |
| [92] | Temperature regime (climate) Extreme temperatures |
Birth weight (continuous variable) | USA (entire country) (NT), 1974–1978 And 1984–1988 |
Number of days during the month or season of birth that fall into different categories of daily average temperature: <20° F,<25° F, 25°–45° F, 45°–65° F, 65°–85° F, >85° F, 90° F, >95° F Winter (December to February), Spring (March to May), Summer (June to August), Fall (September to November) -study period+ annual average temperature from 1960 to 1969 (“climate”) |
Ecological study (county-level resolution) | 20% sample of White mothers aged 19 to 38 | Multilevel linear regression with spatial autocorrelation terms | 4,921,561 | The warmer the yearly average temperature of a county, the lower the birth weight. After controlling for these climatic patterns, birth weight was inversely related to both extremely cold and extremely hot temperatures. In birth month (1974–1978): birth reduction associated with each day <20° F: −0.0761 (SD: 0.0734) >90° F: −0.7449 (SD: 0.0802) With mean county temperature: −1.1409 (SD: 0.3683) In birth month (1984−1988): birth reduction associated with each day <20° F: −0.4749 (SD: 0.0739) >90° F: −0.2927 (SD: 0.06147) With mean county temperature: −4.7054 (SD: 0.2594) |
County per capita income, average elevation |
| [53] | Seasonality of birth (month) |
Birth weight (continuous variable) | Japan (entire country) (NT), January 1974–December 1983 | Year divided into spring (March–May), summer (June–August), autumn (September–November), and winter (December–February) | Time series analysis | All live singletons | Time series regression | N = 16,796,415 | Significant inter-seasonal variability in mean birth weight (p < 0.001): two peaks in May and October–November and two troughs in June–September and December. | None |
| [50] | Seasonality of conception (month) |
Birth weight (continuous variable) Adjusted or not for gestational age |
NJ (entire state), USA, (1997-2006), New York, NY, USA, (1994–2004) PA (entire state), USA, (2004–2010) |
Month of conception | Retrospective cohort study | Single births with no missing information on gestation length | Cohort study based on comparison between siblings | N = 1,435,213 | Gain of 8–9 additional g for summer conceptions compared with January conceptions (both before and after adjusting for gestational age) | Stable maternal characteristics (by design) Influenza Gestational age |
| [82] | Seasonality of birth (Month) Temperature |
Birth weight (continuous variable) | Aberdeen, Scotland (NT), 1950–1956 | -Year divided into Winter (December–February); Spring (March–May); Summer (June–August and Autumn (September–November) -Mean ambient minimum and maximum temperature for 10 days around conception, the middle of Each pregnancy trimester |
Population based cohort study | All births | Linear regression models | N = 12,150 | -lowest birth weights in the winter months (December–February) and highest in the autumn months (September–November) -1 °C increase in mean ambient outdoor temperature in the mid 10-day period of the first trimester -first trimester associated with a 5.4 g (95% CI 2.9, 7.9 g) decrease in birth weight -second trimester associated with a 1.8 g (95% CI −0.7, 4.3 g) decrease in birth weight -third trimester associated with a 1.3 g (95% CI 0.50, 2.1 g) increase in birth weight |
Sex, maternal age, birth year, birth order, social class |
| [96] | Sunlight | Birth weight (continuous variable) | Dunedin, New Zealand (NT), August 1967–July 1978 | -Daily sunlight maximal hours during pregnancy | Hospital based cohort study | All singleton live births | Cross-correlation functions from Fourier transforms | N = 20,021 | - Monthly means for neonate weight varied sinusoidally with monthly variation in mean bright sunlight hours - effect of mean sunlight hours on birth weight most evident when maximal sunlight was positive during the first 3 pre-natal months and negative during the last 6 pre-natal months. |
None |
| [84] | Seasonality of birth (4 seasons) |
Birth weight (continuous variable) | Queensland, Australia (T), January 1987–December 1999 | Year divided into spring (September–November); summer(December–February); fall (March–May); winter(June–August) | Time series analysis | All singleton pregnancies with a gestation of at least 37 weeks | Spectral analysis | N = 350,171 | Winter and spring infants born slightly heavier compared to summer and autumn born infants (25-g difference between neonates born in October vs. May). | None |
| [94] | Temperature Sunlight | Term Birth weight (continuous variable) | Dunedin, New Zealand (NT), January 1999–December 2003 | Temperature and sunshine hours by pregnancy trimester | Hospital based cohort study | Full term births >38 weeks of gestation | One-way analyses of variance | N = 8,516 | - No association between birth weight and temperature in the second trimester. -Infants exposed to high levels of sunshine during the first trimester born heavier than infants exposed to low levels of sunshine. - Infants whose mothers were exposed to trough periods of sunshine during their second and third trimesters heavier than infants whose mothers who were exposed to peak periods of sunshine during the same trimesters |
None |
| [56] | Seasonality of birth (month) |
Birth weight (continuous variable) Low birth weight (<2,500 g) |
USA (entire country) (NT), 1989–2001 | Year divided into months | Population-based cohort study | All birth certificates | Linear regression | 52,041,052 | Children born in December and January have lower average birth weights than other children Infants born in April weigh 23.3 grams more on average than those born in January Early spring and late summer births are less likely to have a low birth weight |
|
| [83] | Seasonality of birth (month) Temperature Precipitation |
Birth weight (continuous variable) Low birth weight (<2,500 g) High birth weight (macrosomia >4,000 g) |
Israel (entire country) (NT), 1998–2004 | -Year divided into seasons: winter (December–February), spring (March–May), summer (June–August), and fall (September–November) - monthly means of maximum and minimum daily temperature, precipitation, and number of rainy days |
Population based cohort study | All live births | Linear regression (mean birth weight) Logistic regression (Low birth weight) |
N = 225,545 | - Significant association between birth weights and season with a peak in July and trough in January - No association between low birth weight and seasonality -Babies born in summer had an OR = 1.12, 95% CI (1.07–1.18) for macrosomia compared with winter. - Positive association between mean birth weight and monthly minimal temperatures at the first month of first and third trimesters. - Monthly means of precipitation and number of rainy days not associated with birth weight |
Maternal age, sex, year of birth, maternal diabetes |
| [85] | Seasonality of birth (month) |
Birth weight (continuous variable) Low birth weight (<2,500 g) Very low birth weight (<1,500 g) |
Kimberly, Australia (T), 1981–1993 | Year divided into seasons: very hot summer (January–June) and heavy rainfall from (January–April); Winter (July–December) | Population based cohort study | All singleton live births | Logistic regression analysis: OR of wet season compared to dry season | N = 4,058 | - Mean birth weight varied by month of birth (p = 0.003) -low birth weight more common during the wet season: OR 2.73; 95% CI (2.3–3.67) with the lowest birth weight in March - Increased risk of very low birth weight during the wet season compared with the dry season: OR 2.73; 95% CI (2.3–3.67), but low birth weight not associated with the wet season OR 1.06; 95% CI (0.96–1.17; p = ns) |
None |
| [75] | Seasonality of conception (4 seasons) |
Low birth weight (<2500 g) Small for gestational age (<10th percentile of birth weight for gestational age) |
NC (entire state), USA (NT), 2001–2005 | -Season defined as: winter (December–February), spring (March–May), summer (June–August), and fall (September–November) | Population based cohort study | Singleton first births to non-Hispanic white and black women, excluding births with missing covariate data, congenital anomalies, birth weight <400 g, extremely high or low gestational age, and maternal age >44 years | Linear regression for mean birth weight logistic regression for low birth weight and small for gestational age | N = 188,276 | -Spring and winter conceptions were associated with higher rates of low birth weight for gestational age among statewide births (p < 0.05), as well as among rural county births for the non-Hispanic white group (p < 0.05). - Rates of small for gestational age were lowest among non-Hispanic white group spring conceptions across all North Carolina counties, urban, and rural counties (p < 0.05) |
Maternal age, education level, marital status, smoking status, region of North Carolina, county urbanization |
| [58] | Seasonality of birth | Small for gestational age (<10th percentile of reference standard gestational age) | Keneba, Manduar, and Kantong Kunda, The Gambia (T), (3 villages of the West Kiang District) 1976–2003 | Year divided into agricultural season that revolves around the rainy season (July–November). | Population based cohort study | All live births | Fourier series | N = 1,916 | Incidence of SGA highest at the end of the annual hungry season, from August to December (peaking in November at 30.6%), with a nadir of 12.9% in June. | Thick and thin blood smears obtained from antenatal clinics to measure malarial infection; activity diaries and 24-hour activity recall to assess maternal workload |
| [66] | Seasonality of conception and birth (4 seasons) Temperature |
Term low birth weight (<2,500 g) | Brandenburg, Germany (NT), 2002–2010 Saxony, Germany (NT), 2005–2009 |
-Year divided into four seasons: December to February (winter), March to May: (spring), June to August: (summer), September to November: (autumn) -Daily mean temperature for each trimester of pregnancy |
Time series analysis | All singleton births ≥37 weeks of gestation and with birth weight greater than 200 g | Logistic time series regression Fourier series |
Brandenburg N = 128,604, including 6,242 low birth weight infants Saxony N = 162,913, including 8,034 low birth weight infants |
- Association between low birth weight and conception in Spring in Brandenburg OR = 1.19, 95% CI (1.05–1.35) - Association between low birth weight and birth in Winter in Brandenburg OR = 1.15, 95% CI (1.02–1.30) - No association between low birth weight and temperature in Brandenburg in first OR = 0.93, 95% CI (0.70–1.23), second OR = 0.91, 95% CI (0.66–2.25), or third trimester OR = 0.86, 95% CI (0.64–1.17) - No association between low birth weight and temperature in Saxony in first OR = 0.89, 95% CI (0.70–1.12), second OR = 1.09, 95% CI (0.82–1.45), or third trimester OR = 1.15, 95% CI (0.87–1.52) |
Maternal age available for Saxony only |
| [95] | Temperature | Very low birth weight (<1,500 g) | Sweden (entire country), (NT), 1973–2010 | Mean daily temperature averaged for the month of birth | Population based cohort study | All singleton live births during the summer season (June, July, August) | Time series analysis | N = 3,757,440 | - Inverse association between very low birth weight risk and mean monthly temperature in summer season - 13.6% increase in odds of a very low birth weight male for a colder than expected June and 5.4% increase in odds for a colder than expected August |
|
| [32] | Seasonality of birth (rainy vs. dry) |
Low birth weight (<2,500 g) Small-for-gestation age (birth weight <10th centile of the gestational age- and sex-specific US reference for fetal growth) |
Lombok, Indonesia (T), 2001–2004 | Year divided into rainy season (November–March) and dry season (April–October) | Double blind cluster randomized controlled trial | All singleton live births with birth weight measured within 72 h of birth | Hierarchical logistic regression | N = 14,040 | 22% increased odds of low birth weight in babies born in the rainy season; 18% increased odds of small for gestational age in babies born in the rainy season |
Infant’s sex, season at birth, mothers’ residence, nutritional status, education, household wealth, mid-upper arm circumference, height and a composite variable of birth order and pregnancy interval |
| [88] | Seasonality of birth (4 seasons) Temperature Sunlight |
Small for gestational age and sex (infants with a weight for gestational age <10th percentile for their sex) Proportion of optimal birth weight (POBW) | Perth, Australia (NT), 1998–2006 |
-Year divided into seasons: winter (June–August) and summer (December–February) -temperature and sunlight averaged over the entire duration pregnancy and over each trimester of pregnancy separately. |
Population based cohort study | All singleton live births ≥400 g birth weight and/or ≥20 weeks’ gestation |
Multiple linear regression, multivariate models | N = 14,7357 | - POBW with third trimesters predominantly in summer was 0.18%, 95% CI (0.00%–0.36%) lower than for those in winter. - No association between season of birth and small for gestational age - Inter-quartile range increase in temperature during entire pregnancy (0.73 °C) was associated with small for gestational age and sex with an OR = 1.02, 95% CI (1.00–1.05). - POBW decreased by 0.14%, 95% CI (0.01%–0.27%) per inter-quartile range increase in third-trimester temperature (9.15 °C). - No significant effect observed for sunlight exposure |
Criteria air pollutants: Particulate matter with aerodynamic diameter <2.5 micrometers and <10 micrometers, ozone, nitric oxide, nitrogen dioxide and carbon monoxide |
a NT: Non-tropical climate; T: Tropical climate.