The Effect of the Menstrual Cycle on Energy Intake: A Systematic Review and Meta-analysis

PECOS Characteristics and Key Words in the Search Strategy for Energy Intake Across the Menstrual Cycle

PECOS guideline	Definition	Key words search
Population	Regularly menstruating females	Menstrua, ovarian, ovulatory
Exposure	Any intervention	Luteal
Comparison	Comparator to exposure	Follicular
Outcome	Energy intake across the menstrual cycle	Energy intake, food intake, nutrient intake, calor*
Study design	Any primary studies	Not set

PECOS guideline	Definition	Key words search
Population	Regularly menstruating females	Menstrua, ovarian, ovulatory
Exposure	Any intervention	Luteal
Comparison	Comparator to exposure	Follicular
Outcome	Energy intake across the menstrual cycle	Energy intake, food intake, nutrient intake, calor*
Study design	Any primary studies	Not set

Table 1.

PECOS Characteristics and Key Words in the Search Strategy for Energy Intake Across the Menstrual Cycle

PECOS guideline	Definition	Key words search
Population	Regularly menstruating females	Menstrua, ovarian, ovulatory
Exposure	Any intervention	Luteal
Comparison	Comparator to exposure	Follicular
Outcome	Energy intake across the menstrual cycle	Energy intake, food intake, nutrient intake, calor*
Study design	Any primary studies	Not set

PECOS guideline	Definition	Key words search
Population	Regularly menstruating females	Menstrua, ovarian, ovulatory
Exposure	Any intervention	Luteal
Comparison	Comparator to exposure	Follicular
Outcome	Energy intake across the menstrual cycle	Energy intake, food intake, nutrient intake, calor*
Study design	Any primary studies	Not set

Study selection followed a 2-step process and was independently completed by 2 reviewers (J.A.L.T. and J.S.K.). A blinded third reviewer (D.P.D.B.) settled any discrepancies independently. Titles and abstracts of articles were first screened to filter out studies that were irrelevant and did not relate to the study purpose in the Covidence software. Relevancy was determined individually by the reviewers and each article was considered relevant if the title and abstract discussed the following key words (a process assisted by an automated system within Covidence); (1) follicular, (2) luteal, (3) ovulatory, (4) eumenorrheic, (5) normally menstruating, (6) energy intake, (7) food intake, and (8) dietary intake.

Assessment of study quality

Three reviewers (J.A.L.T., D.P.D.B., and S.F.M.) independently assessed the methodological quality and risk of bias in each study using a modified tool previously developed and utilized by authors for situations in which tools such as Cochrane’s Risk of Bias tool was not appropriate for their objective.³⁷ This modified tool was developed to account for limitations in sample size and inclusion criteria, measurement of outcome variables, and control of confounding variables. All studies included in analysis were identified to be either low, moderate, or high quality in 3 key categories: (1) sample size and inclusion criteria, (2) measurement of outcome, and (3) control for confounders, as described previously.³⁷ All 3 reviewers completed the assessment independently, and subsequently met on February 14, 2024, to resolve any discrepancies between them. For each discrepancy, if 2 of the 3 reviewers agreed on a decision (to either include or not include the study), the decision agreed upon by the majority was determined. In the case that all 3 reviewers disagreed, each reviewer presented their individual case and a fourth author (T.J.H.) made the final determination (Table 2). The presence of a sample size calculation and clear statement of inclusion and exclusion criteria was evaluated and each study was determined to be of low (no sample size calculation or statement of inclusion criteria), moderate (no sample size calculation and lack of clarity on inclusion and exclusion criteria), or high (clear statement of sample size calculation and both inclusion and exclusion criteria) quality. For the evaluation of measurement outcomes, each study was evaluated on the presence of efficacious menstrual cycle determination and measurement of energy intake, and each study was determined to be of low (cycle tracking as the only use of menstrual cycle determination as well as reliance on retrospective methods for energy intake measurement), moderate (>1 method of menstrual cycle determination that includes cycle tracking in addition to serum hormone concentrations or the use of LH strips, and the use of retrospective energy intake methodology), or high (>1 method of menstrual cycle determination that includes cycle tracking in addition to serum hormone concentrations or the use of LH strips, and the use of prospective energy intake methodology) quality. For the evaluation of control for confounding variables, each study was evaluated individually on how clearly it controlled for caffeine intake, presence of smokers, medication use, and presence of potential comorbidities and each was assigned a ranking of low (no accountability or control for any of the possible confounding factors), moderate (accounting for 1 or 2 of the listed confounding factors), or high (accounting for all of the listed confounding factors) quality accordingly.

Table 2.

Quality Assessment Analysis Across All Studies Included in the Analysis

Study, year	Sample size and eligibility criteria	Measurement of outcome	Control for confounders
Barr et al, 1995⁴³	Moderate	Moderate	Moderate
Chappell and Hackney, 1997²⁰	Moderate	Moderate	High
Cheikh Ismail et al, 2009¹⁹	Moderate	Moderate	Moderate
Souza et al, 2018³¹	Moderate	Moderate	Moderate
Eck et al, 1997⁴²	Moderate	Moderate	High
Elliott et al, 2015¹⁰	Moderate	Moderate	Moderate
Fong and Kretsch, 1993²⁹	Moderate	Moderate	Moderate
Gil et al, 2009²¹	Moderate	High	High
Gong et al, 1989²²	Low	Moderate	Moderate
Ihalainen et al, 2021²⁷	Low	High	Moderate
Johnson et al, 1993⁴¹	Moderate	High	Moderate
Kammoun et al, 2016²³	Moderate	Moderate	Moderate
Li et al, 1999²⁴	Moderate	Moderate	Low
Lyons et al, 1989²⁵	Moderate	Moderate	Moderate
Martini et al, 1994²⁶	Moderate	High	High

Study, year	Sample size and eligibility criteria	Measurement of outcome	Control for confounders
Barr et al, 1995⁴³	Moderate	Moderate	Moderate
Chappell and Hackney, 1997²⁰	Moderate	Moderate	High
Cheikh Ismail et al, 2009¹⁹	Moderate	Moderate	Moderate
Souza et al, 2018³¹	Moderate	Moderate	Moderate
Eck et al, 1997⁴²	Moderate	Moderate	High
Elliott et al, 2015¹⁰	Moderate	Moderate	Moderate
Fong and Kretsch, 1993²⁹	Moderate	Moderate	Moderate
Gil et al, 2009²¹	Moderate	High	High
Gong et al, 1989²²	Low	Moderate	Moderate
Ihalainen et al, 2021²⁷	Low	High	Moderate
Johnson et al, 1993⁴¹	Moderate	High	Moderate
Kammoun et al, 2016²³	Moderate	Moderate	Moderate
Li et al, 1999²⁴	Moderate	Moderate	Low
Lyons et al, 1989²⁵	Moderate	Moderate	Moderate
Martini et al, 1994²⁶	Moderate	High	High

See Benton et al, 2020³⁷. “High” indicates high quality and therefore papers exhibit a low risk of bias; “Moderate” indicates moderate quality and therefore papers exhibit a moderate risk of bias; “Low” indicates low quality and therefore papers exhibit a high risk of bias.

Table 2.

Open in new tab Download slide

Quality Assessment Analysis Across All Studies Included in the Analysis

Study, year	Sample size and eligibility criteria	Measurement of outcome	Control for confounders
Barr et al, 1995⁴³	Moderate	Moderate	Moderate
Chappell and Hackney, 1997²⁰	Moderate	Moderate	High
Cheikh Ismail et al, 2009¹⁹	Moderate	Moderate	Moderate
Souza et al, 2018³¹	Moderate	Moderate	Moderate
Eck et al, 1997⁴²	Moderate	Moderate	High
Elliott et al, 2015¹⁰	Moderate	Moderate	Moderate
Fong and Kretsch, 1993²⁹	Moderate	Moderate	Moderate
Gil et al, 2009²¹	Moderate	High	High
Gong et al, 1989²²	Low	Moderate	Moderate
Ihalainen et al, 2021²⁷	Low	High	Moderate
Johnson et al, 1993⁴¹	Moderate	High	Moderate
Kammoun et al, 2016²³	Moderate	Moderate	Moderate
Li et al, 1999²⁴	Moderate	Moderate	Low
Lyons et al, 1989²⁵	Moderate	Moderate	Moderate
Martini et al, 1994²⁶	Moderate	High	High

Study, year	Sample size and eligibility criteria	Measurement of outcome	Control for confounders
Barr et al, 1995⁴³	Moderate	Moderate	Moderate
Chappell and Hackney, 1997²⁰	Moderate	Moderate	High
Cheikh Ismail et al, 2009¹⁹	Moderate	Moderate	Moderate
Souza et al, 2018³¹	Moderate	Moderate	Moderate
Eck et al, 1997⁴²	Moderate	Moderate	High
Elliott et al, 2015¹⁰	Moderate	Moderate	Moderate
Fong and Kretsch, 1993²⁹	Moderate	Moderate	Moderate
Gil et al, 2009²¹	Moderate	High	High
Gong et al, 1989²²	Low	Moderate	Moderate
Ihalainen et al, 2021²⁷	Low	High	Moderate
Johnson et al, 1993⁴¹	Moderate	High	Moderate
Kammoun et al, 2016²³	Moderate	Moderate	Moderate
Li et al, 1999²⁴	Moderate	Moderate	Low
Lyons et al, 1989²⁵	Moderate	Moderate	Moderate
Martini et al, 1994²⁶	Moderate	High	High

Data extraction

The following participant characteristics were extracted: mean age (years), mean BMI (kg/m²), and mean length of menstrual cycle (days). The following study characteristics were extracted: menstrual cycle definition, approximate days in the cycle when energy intake was measured, number of menstrual cycles studied, which menstrual phases were studied, methods for determining menstrual phase, how long energy intake was tracked, how energy intake was measured, if energy intake was measured consecutively or sporadically in each phase, and absolute energy intake in each phase. The means and SDs for all data were extracted by recording all values directly from each article included in this meta-analysis. Martini et al²⁶ provided SEs, which were converted to SDs.

Statistical analysis

Across studies, energy intake was presented in kilocalories per day (kcal⋅d⁻¹), kilojoules per day (kJ⋅d⁻¹), or megajoules per day (MJ⋅d⁻¹). All means and SDs for energy intake were converted to kilocalories per day (1 kcal = 4.18 kJ = 0.00418 MJ). To improve clinical interpretation, a crude difference in energy intake as kcal⋅d⁻¹ was calculated between the phases for each study included in this analysis. Mean energy intakes in the FP and LP were first averaged across the studies, and subsequently the average energy intake in the FP was subtracted from the average energy intake in the LP. It is important to note that this calculation does not account for the varying weights the individual studies received in the actual meta-analysis. Statistical analysis was conducted using R (version 2023.12.1 + 402) in R-Studio (version 2023.12.1 + 402; Posit Software, Boston, United States of America) and we used the following resource³⁹ to complete the analysis. Specifically, we used section 3.3.1.3 to calculate the standardized mean difference (SMD) (eqn 1) and the SE (eqn 2) for each study, and section 4.2.1 to guide the meta-analysis. The random-effects meta-analysis was conducted using the metagen function from the R package meta t.³⁹

To calculate SMD, energy intake during the FP ${(\bar{x}}_{F P})$ was subtracted from energy intake during the LP (⁠ ${\bar{x}}_{L P})$ and divided by the SD during the FP (SD_FP). Harrer et al³⁹ suggests using the SD from the initial test (assuming a pre-post design) when conducting a within-subject analysis; however, as not all studies had participants complete the sessions in the same order, we used the FP data for all calculations. The average difference in SD between phases was only 10 kcal and when SMD was calculated with SD_LP and the model run with these data, the overall effect was still statistically significant (data not shown).

SMD = \frac{{\bar{x}}_{LP} - {\bar{x}}_{FP}}{{SD}_{F P}^{}}

(1)

To calculate the SE (eqn 2) the SMD value (SMD; previously calculated), along with the number of participants (n) and the correlation between the 2 phases were used. As no published research had correlated energy intake during the FP and LP, and none of the corresponding authors of the studies included in this manuscript returned any e-mails regarding correlation values from their respective studies, a conservative r value was used (0.2, low correlation) for all SE calculations.

SE = \sqrt{\frac{2 (1 - r_{t 1 t 2})}{n} + \frac{{SMD}^{2}}{2 n}}

(2)

RESULTS

Study selection

A comprehensive literature search was conducted in PubMed, Medline, CINAHL, Embase, and Web of Science on January 6, 2023, and a second up-to-date search took place on January 23, 2023⁴⁰ (Figure 1). From this search, a total of 10 037 articles were imported into Covidence. From this, 5203 articles were removed as they were automatically characterized as duplicates. Titles and abstracts of 4834 articles were first screened to filter out studies that were irrelevant and did not relate to the study purpose in the Covidence software. Seventy-six studies passed the title and abstract screening, and the full-text manuscripts were downloaded directly from Covidence and reviewed to determine if the eligibility criteria were met. Of the 76 studies that were included in full-text screening, 19 studies did not measure energy intake for a full day, 11 were considered to be a secondary analysis, 6 contained smokers in the study population, 5 full-texts could not be completely accessed, 5 were conference abstracts and not published papers, 5 provided incomplete datasets, 5 looked at the influence of diets/dieting only, 3 studied participants with food-related disorders, and 2 were animal studies. A total of 15 studies passed both stages of screening and were extracted and included in the final analysis.

Figure 1.

Flowchart of the Study Selection Process and Studies Included in the Meta-Analysis. Abbreviation: EI, energy intake

Study and participant characteristics

The 15 datasets included in this meta-analysis consisted of 330 female participants (Table 3) with a mean age of 26 ± 4 years (range, 20–36 years) and mean BMI of 22.4 ± 2.3 kg/m² (range, 18.7–26.9 kg/m²). All studies, with the exception of 2, recruited participants between the ages of 20 and 30 years.²²^,⁴¹ Only 4 studies clearly identified that the population recruited was recreationally active,²⁰^,²⁶^,²⁷^,⁴¹ while the other studies provided no definition of physical activity status. Four studies included in this analysis involved different datasets or groups within the study, which remained separated, and these specific datasets were not used.²⁰^,²⁷^,⁴² One of these studies involved the comparison of physically active females to physically inactive females²⁰ and 2 studies compared users of monophasic hormonal birth control with those who were not.²⁷^,⁴²

Table 3.

Study Characteristics Outlining Participant Characteristics and Methodology for Energy Intake and Menstrual Cycle

Study, year	Age (y) and BMI (kg/m²)	MC length (d)	MC definition		Cycles studied	Method	Energy intake		No. of days EI measured (d·phase)	Average EI across each phase, kcal·d^-1 (kJ·d^-1)
Study, year	Age (y) and BMI (kg/m²)	MC length (d)	FP	LP	Cycles studied	Method	FP	LP	No. of days EI measured (d·phase)	FP	LP
Barr et al, 1995⁴³	28 ± 5	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	1915 ± 565	2216 ± 642
Barr et al, 1995⁴³	21.8 ± 2.1	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	(8012 ± 2363)	(9271 ± 2686)
Chappell and Hackney, 2007²⁰	22 ± 1	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	1901 ± 124	2037 ± 150
Chappell and Hackney, 2007²⁰	21.3	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	(7953 ± 518)	(8522 ± 627)
Cheikh Ismail et al, 2009¹⁹	22 ± 3	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	1255 ± 526	1363 ± 550
Cheikh Ismail et al, 2009¹⁹	22.2 ± 3.9	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	(5250 ± 2200)	(5702 ± 2301)
Eck et al, 1997⁴²	20 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	1601 ± 359	1769 ± 359
Eck et al, 1997⁴²	20.8 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	(6698 ± 1502)	(7401 ± 1502)
Elliott et al, 2015¹⁰	24 ± 1	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	1663 ± 413	1384 ± 307
Elliott et al, 2015¹⁰	20.2 ± 2.8	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	(6957 ± 1727)	(5790 ± 1284)
Fong and Kretsch, 1993²⁹	28 ± 4	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	2360 ± 543	2501 ± 539
Fong and Kretsch, 1993²⁹	22.4 ± 2.0	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	(9874 ± 2271)	(10464 ± 2255)
Gil et al, 2009²¹	28 ± 4	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	1705 ± 251	2227 ± 370
Gil et al, 2009²¹	22.20 ± 2.0	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	(7133 ± 1050)	(9317 ± 1548)
Gong et al, 1989²²	31 ± 7	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	1833 ± 146	2040 ± 156
Gong et al, 1989²²	22.4	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	(7669 ± 610)	(8535 ± 652)
Ihalainen et al, 2021²⁷	26 ± 4	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	2340 ± 540	2270 ± 370
Ihalainen et al, 2021²⁷	24.0	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	(9790 ± 2259)	(2497 ± 1548)
Johnson et al, 1993⁴¹	32 ± 4	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	1711 ± 421	1875 ± 445
Johnson et al, 1993⁴¹	21.5	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	(7158 ± 1761)	(7845 ± 1861)
Kammoun et al, 2016²³	27 ± 8	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	1688 ± 332	2164 ± 322
Kammoun et al, 2016²³	25.6 ± 5.2	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	(7062 ± 1389)	(9054 ± 1347)
Li et al, 1999²⁴	21 ± 1	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	1457 ± 281	1668 ± 441
Li et al, 1999²⁴	19.6 ± 1.4	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	(6096 ± 1175)	(7062 ± 1845)
Lyons et al, 1989²⁵	24	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	2012 ± 69	2150 ± 86
Lyons et al, 1989²⁵	22.8	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	(8418 ± 288)	(8995 ± 359)
Martini et al, 1994²⁶	27 ± 4	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	1749 ± 37	1908 ± 38
Martini et al, 1994²⁶	22.1 ± 1.9	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	(7319 ± 154)	(7983 ± 158)
Souza et al, 2018³¹	22 ± 1	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	1694 ± 437	1738 ± 414
Souza et al, 2018³¹	23.4 ± 0.9	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	(7087 ± 1828)	(7271 ± 1732)
Pooled characteristics	26 ± 4	29 ± 2								1791 ± 307	1959 ± 346
Pooled characteristics	22.4 ± 2.3	29 ± 2								(7494 ± 1284)	(8196 ± 1445)

Study, year	Age (y) and BMI (kg/m²)	MC length (d)	MC definition		Cycles studied	Method	Energy intake		No. of days EI measured (d·phase)	Average EI across each phase, kcal·d^-1 (kJ·d^-1)
Study, year	Age (y) and BMI (kg/m²)	MC length (d)	FP	LP	Cycles studied	Method	FP	LP	No. of days EI measured (d·phase)	FP	LP
Barr et al, 1995⁴³	28 ± 5	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	1915 ± 565	2216 ± 642
Barr et al, 1995⁴³	21.8 ± 2.1	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	(8012 ± 2363)	(9271 ± 2686)
Chappell and Hackney, 2007²⁰	22 ± 1	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	1901 ± 124	2037 ± 150
Chappell and Hackney, 2007²⁰	21.3	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	(7953 ± 518)	(8522 ± 627)
Cheikh Ismail et al, 2009¹⁹	22 ± 3	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	1255 ± 526	1363 ± 550
Cheikh Ismail et al, 2009¹⁹	22.2 ± 3.9	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	(5250 ± 2200)	(5702 ± 2301)
Eck et al, 1997⁴²	20 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	1601 ± 359	1769 ± 359
Eck et al, 1997⁴²	20.8 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	(6698 ± 1502)	(7401 ± 1502)
Elliott et al, 2015¹⁰	24 ± 1	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	1663 ± 413	1384 ± 307
Elliott et al, 2015¹⁰	20.2 ± 2.8	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	(6957 ± 1727)	(5790 ± 1284)
Fong and Kretsch, 1993²⁹	28 ± 4	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	2360 ± 543	2501 ± 539
Fong and Kretsch, 1993²⁹	22.4 ± 2.0	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	(9874 ± 2271)	(10464 ± 2255)
Gil et al, 2009²¹	28 ± 4	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	1705 ± 251	2227 ± 370
Gil et al, 2009²¹	22.20 ± 2.0	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	(7133 ± 1050)	(9317 ± 1548)
Gong et al, 1989²²	31 ± 7	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	1833 ± 146	2040 ± 156
Gong et al, 1989²²	22.4	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	(7669 ± 610)	(8535 ± 652)
Ihalainen et al, 2021²⁷	26 ± 4	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	2340 ± 540	2270 ± 370
Ihalainen et al, 2021²⁷	24.0	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	(9790 ± 2259)	(2497 ± 1548)
Johnson et al, 1993⁴¹	32 ± 4	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	1711 ± 421	1875 ± 445
Johnson et al, 1993⁴¹	21.5	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	(7158 ± 1761)	(7845 ± 1861)
Kammoun et al, 2016²³	27 ± 8	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	1688 ± 332	2164 ± 322
Kammoun et al, 2016²³	25.6 ± 5.2	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	(7062 ± 1389)	(9054 ± 1347)
Li et al, 1999²⁴	21 ± 1	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	1457 ± 281	1668 ± 441
Li et al, 1999²⁴	19.6 ± 1.4	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	(6096 ± 1175)	(7062 ± 1845)
Lyons et al, 1989²⁵	24	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	2012 ± 69	2150 ± 86
Lyons et al, 1989²⁵	22.8	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	(8418 ± 288)	(8995 ± 359)
Martini et al, 1994²⁶	27 ± 4	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	1749 ± 37	1908 ± 38
Martini et al, 1994²⁶	22.1 ± 1.9	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	(7319 ± 154)	(7983 ± 158)
Souza et al, 2018³¹	22 ± 1	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	1694 ± 437	1738 ± 414
Souza et al, 2018³¹	23.4 ± 0.9	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	(7087 ± 1828)	(7271 ± 1732)
Pooled characteristics	26 ± 4	29 ± 2								1791 ± 307	1959 ± 346
Pooled characteristics	22.4 ± 2.3	29 ± 2								(7494 ± 1284)	(8196 ± 1445)

Represents percentage of cycle length for determining menstrual phase.

Abbreviations: BMI, body mass index; E₂, estradiol; [E₂/P₄], serum E₂/P₄ concentrations; EI, energy intake; FP, follicular phase; LH, luteinizing hormone; LP, luteal phase; MC, menstrual cycle; NA, not available; P₄, progesterone; Temp, temperature; +ve, positive.

Table 3.

Open in new tab Download slide

Study Characteristics Outlining Participant Characteristics and Methodology for Energy Intake and Menstrual Cycle

Study, year	Age (y) and BMI (kg/m²)	MC length (d)	MC definition		Cycles studied	Method	Energy intake		No. of days EI measured (d·phase)	Average EI across each phase, kcal·d^-1 (kJ·d^-1)
Study, year	Age (y) and BMI (kg/m²)	MC length (d)	FP	LP	Cycles studied	Method	FP	LP	No. of days EI measured (d·phase)	FP	LP
Barr et al, 1995⁴³	28 ± 5	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	1915 ± 565	2216 ± 642
Barr et al, 1995⁴³	21.8 ± 2.1	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	(8012 ± 2363)	(9271 ± 2686)
Chappell and Hackney, 2007²⁰	22 ± 1	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	1901 ± 124	2037 ± 150
Chappell and Hackney, 2007²⁰	21.3	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	(7953 ± 518)	(8522 ± 627)
Cheikh Ismail et al, 2009¹⁹	22 ± 3	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	1255 ± 526	1363 ± 550
Cheikh Ismail et al, 2009¹⁹	22.2 ± 3.9	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	(5250 ± 2200)	(5702 ± 2301)
Eck et al, 1997⁴²	20 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	1601 ± 359	1769 ± 359
Eck et al, 1997⁴²	20.8 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	(6698 ± 1502)	(7401 ± 1502)
Elliott et al, 2015¹⁰	24 ± 1	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	1663 ± 413	1384 ± 307
Elliott et al, 2015¹⁰	20.2 ± 2.8	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	(6957 ± 1727)	(5790 ± 1284)
Fong and Kretsch, 1993²⁹	28 ± 4	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	2360 ± 543	2501 ± 539
Fong and Kretsch, 1993²⁹	22.4 ± 2.0	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	(9874 ± 2271)	(10464 ± 2255)
Gil et al, 2009²¹	28 ± 4	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	1705 ± 251	2227 ± 370
Gil et al, 2009²¹	22.20 ± 2.0	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	(7133 ± 1050)	(9317 ± 1548)
Gong et al, 1989²²	31 ± 7	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	1833 ± 146	2040 ± 156
Gong et al, 1989²²	22.4	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	(7669 ± 610)	(8535 ± 652)
Ihalainen et al, 2021²⁷	26 ± 4	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	2340 ± 540	2270 ± 370
Ihalainen et al, 2021²⁷	24.0	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	(9790 ± 2259)	(2497 ± 1548)
Johnson et al, 1993⁴¹	32 ± 4	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	1711 ± 421	1875 ± 445
Johnson et al, 1993⁴¹	21.5	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	(7158 ± 1761)	(7845 ± 1861)
Kammoun et al, 2016²³	27 ± 8	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	1688 ± 332	2164 ± 322
Kammoun et al, 2016²³	25.6 ± 5.2	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	(7062 ± 1389)	(9054 ± 1347)
Li et al, 1999²⁴	21 ± 1	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	1457 ± 281	1668 ± 441
Li et al, 1999²⁴	19.6 ± 1.4	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	(6096 ± 1175)	(7062 ± 1845)
Lyons et al, 1989²⁵	24	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	2012 ± 69	2150 ± 86
Lyons et al, 1989²⁵	22.8	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	(8418 ± 288)	(8995 ± 359)
Martini et al, 1994²⁶	27 ± 4	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	1749 ± 37	1908 ± 38
Martini et al, 1994²⁶	22.1 ± 1.9	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	(7319 ± 154)	(7983 ± 158)
Souza et al, 2018³¹	22 ± 1	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	1694 ± 437	1738 ± 414
Souza et al, 2018³¹	23.4 ± 0.9	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	(7087 ± 1828)	(7271 ± 1732)
Pooled characteristics	26 ± 4	29 ± 2								1791 ± 307	1959 ± 346
Pooled characteristics	22.4 ± 2.3	29 ± 2								(7494 ± 1284)	(8196 ± 1445)

Study, year	Age (y) and BMI (kg/m²)	MC length (d)	MC definition		Cycles studied	Method	Energy intake		No. of days EI measured (d·phase)	Average EI across each phase, kcal·d^-1 (kJ·d^-1)
Study, year	Age (y) and BMI (kg/m²)	MC length (d)	FP	LP	Cycles studied	Method	FP	LP	No. of days EI measured (d·phase)	FP	LP
Barr et al, 1995⁴³	28 ± 5	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	1915 ± 565	2216 ± 642
Barr et al, 1995⁴³	21.8 ± 2.1	NA	NA	NA	3	Temp	Day 3-8	Day 20-26	3	(8012 ± 2363)	(9271 ± 2686)
Chappell and Hackney, 2007²⁰	22 ± 1	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	1901 ± 124	2037 ± 150
Chappell and Hackney, 2007²⁰	21.3	30 ± 2	NA	NA	1	Temp	Day 10	Day 20	10	(7953 ± 518)	(8522 ± 627)
Cheikh Ismail et al, 2009¹⁹	22 ± 3	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	1255 ± 526	1363 ± 550
Cheikh Ismail et al, 2009¹⁹	22.2 ± 3.9	29 ± 2	Day 0-10	Day 20–menses	1	Cycle tracking	NA	NA	2	(5250 ± 2200)	(5702 ± 2301)
Eck et al, 1997⁴²	20 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	1601 ± 359	1769 ± 359
Eck et al, 1997⁴²	20.8 ± 3	27 ± 2	Day 0-13	Day 15–menses	2	LH kits	Day 7-9	Day 22-24	Daily	(6698 ± 1502)	(7401 ± 1502)
Elliott et al, 2015¹⁰	24 ± 1	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	1663 ± 413	1384 ± 307
Elliott et al, 2015¹⁰	20.2 ± 2.8	30 ± 6	25-49%^a	51-100%^a	1 or 2	Cycle tracking	NA	NA	3	(6957 ± 1727)	(5790 ± 1284)
Fong and Kretsch, 1993²⁹	28 ± 4	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	2360 ± 543	2501 ± 539
Fong and Kretsch, 1993²⁹	22.4 ± 2.0	26 ± 2	Day 4-11	Day 16-28	1	Temp	NA	NA	Daily	(9874 ± 2271)	(10464 ± 2255)
Gil et al, 2009²¹	28 ± 4	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	1705 ± 251	2227 ± 370
Gil et al, 2009²¹	22.20 ± 2.0	NA	Day 5-9	Day 20-28	1	[E₂/P₄]	NA	NA	3	(7133 ± 1050)	(9317 ± 1548)
Gong et al, 1989²²	31 ± 7	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	1833 ± 146	2040 ± 156
Gong et al, 1989²²	22.4	27 ± 2	Day 6-14	Day 15-18	1	Temp	Day 5-13	Day 15-28	Daily	(7669 ± 610)	(8535 ± 652)
Ihalainen et al, 2021²⁷	26 ± 4	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	2340 ± 540	2270 ± 370
Ihalainen et al, 2021²⁷	24.0	NA	NA	NA	1	LH kits and [E₂/P₄]	Day 7-11	7 Days post-LH surge	3	(9790 ± 2259)	(2497 ± 1548)
Johnson et al, 1993⁴¹	32 ± 4	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	1711 ± 421	1875 ± 445
Johnson et al, 1993⁴¹	21.5	28 ± 5	Rising E₂, low P₄	Peak P₄, increase in E₂	1	Temp and [E₂/P₄]	NA	NA	Daily	(7158 ± 1761)	(7845 ± 1861)
Kammoun et al, 2016²³	27 ± 8	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	1688 ± 332	2164 ± 322
Kammoun et al, 2016²³	25.6 ± 5.2	NA	NA	NA	1	Cycle tracking	Day 1-3	Day 17-25	1	(7062 ± 1389)	(9054 ± 1347)
Li et al, 1999²⁴	21 ± 1	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	1457 ± 281	1668 ± 441
Li et al, 1999²⁴	19.6 ± 1.4	31 ± 2	Day 0-10	6-10 Days post-LH surge	1	LH kits	NA	NA	3	(6096 ± 1175)	(7062 ± 1845)
Lyons et al, 1989²⁵	24	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	2012 ± 69	2150 ± 86
Lyons et al, 1989²⁵	22.8	28 ± 1	Day 5 – +ve LH test	4 Days post-LH surge	1	LH kits	NA	NA	Daily	(8418 ± 288)	(8995 ± 359)
Martini et al, 1994²⁶	27 ± 4	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	1749 ± 37	1908 ± 38
Martini et al, 1994²⁶	22.1 ± 1.9	NA	Day 0-9	7-9 Days post-LH surge	4, 5, or 6	LH kits, [E₂/P₄], temp	Day 7-9	7-9 Days post-LH surge	3	(7319 ± 154)	(7983 ± 158)
Souza et al, 2018³¹	22 ± 1	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	1694 ± 437	1738 ± 414
Souza et al, 2018³¹	23.4 ± 0.9	NA	NA	NA	1	Cycle tracking	Day 5-9	Day 20-25	1	(7087 ± 1828)	(7271 ± 1732)
Pooled characteristics	26 ± 4	29 ± 2								1791 ± 307	1959 ± 346
Pooled characteristics	22.4 ± 2.3	29 ± 2								(7494 ± 1284)	(8196 ± 1445)

Represents percentage of cycle length for determining menstrual phase.

Study quality

The assessment of study quality is shown in Table 2. For the evaluation of sample size and inclusion and exclusion criteria, all studies but 2 were considered as having moderate quality,²²^,²⁷ where neither a sample size calculation nor clear indications of inclusion and exclusion criteria were present. For measurement outcomes, all studies were considered of moderate quality with the exception of 4 studies,²¹^,²⁶^,²⁷^,⁴¹ which exhibited high-quality measurements of energy intake and menstrual cycle estimation. Finally, all studies but 4 were considered high quality in terms of controlling for possible confounding variables,²⁰^,²¹^,²⁶^,⁴² and 1 study was considered to be low quality²⁴ for a lack of addressing possible confounders directly.

Difference in energy intake between menstrual cycle phases

The overall effect size was moderate (SMD = 0.69, P = .039, n = 330), and the interstudy heterogeneity was estimated to be τ² = 0.4972, with an I² = 83% (P < .01).³⁸ A forest plot (Figure 2) presenting the differences in energy intake between the FP and LP was generated. An overview of the energy intake data in individual studies is presented in Table 3. This plot includes each individual study, the corresponding effect sizes calculated using SMD, the weighted percentage, and the overall effect size. The results identified an average increase in energy intake in the LP compared with the FP, with an average difference of 168 kcal⋅d⁻¹.

Figure 2.

Forest Plot of Individual Effect Sizes (Calculated Using SMD), Weighted Percentage, and Overall Effect Size Comparing the Difference in Energy Intake Between the Follicular Phase and Luteal Phase. Positive values favor greater energy intake in the LP compared with the FP. Abbreviations: FP, follicular phase; HK, Knapp-Hartung adjustment; LP, luteal phase; SMD, standardized mean difference

DISCUSSION

While the impact of menstrual cycle phase on energy intake has been explored previously,¹⁰^,^20–22^,^24–27^,²⁹^,³¹^,^41–43 these studies demonstrated conflicting results and have yet to be synthesized. Therefore, this systematic review and meta-analysis investigated if differences in energy intake exist across the menstrual cycle phases that represent the majority of the menstrual cycle.¹⁵ Energy intake was higher in the LP compared with the FP (P = .39, SMD = 0.69) (Figure 2), with an increase in energy intake in the LP compared with the FP by 168 kcal⋅d⁻¹. Importantly, there was a large degree of heterogeneity in the measurement of energy intake as well as menstrual phase determination, although this is to be expected as female-specific research has evolved greatly over the last 4 decades and methodological designs have improved greatly.¹⁵

Of the 15 datasets included in analysis, 8 reported statistical differences in energy intake between the LP and FP¹⁰^,^20–24^,²⁵^,²⁶ and 7 reported no differences.¹⁹^,²⁷^,²⁹^,³¹^,^41–43 Certain studies had multiple datasets to include those with oral contraceptives,²⁷^,⁴² while 1 study had 2 groups in which 1 was considered to be physically inactive²⁰; however, only the groups that were not taking oral contraceptives and were characterized specifically as being recreationally active were used for analysis. Of the 7 studies that reported no statistically significant differences, 6 reported effect sizes that favored an increase in energy intake in the LP compared with the FP (average, ∼99 kcal⋅d⁻¹).¹⁹^,²⁹^,³¹^,^41–43 Of the 8 studies that identified differences in energy intake between FP and LP, 7 demonstrated greater energy intake (average, ∼275 kcal⋅d⁻¹) in the LP compared with the FP.^20–24^,²⁵^,²⁶ Interestingly, only 1 study actually demonstrated greater energy intake in the FP compared with the LP,¹⁰ with a difference of approximately 260 kcal⋅d⁻¹. There is no apparent physiological explanation to support greater energy intake in the FP compared with the LP for this particular study. The discrepancy between these studies could likely be due to the result of methodological differences. This particular study identified a greater energy intake in the FP and used a percentage of cycle length completed to determine the phase in which menstruation marked 0% and the FP and LP were 25–49% and 50–100%, respectively. Menstrual phase was not confirmed via LH strips or blood sampling,¹⁰ which has now been identified as a critical step in accurate menstrual phase characterization.¹⁵ This meta-analysis as well as the majority of research support greater energy intake in the LP compared with the FP, which is line with both murine model data as well as the hypothesis that a greater energy intake in the LP is a biological response to support a potential pregnancy, by ensuring that females are not in an energy deficit during a period of high energy demand.^45–47 There was large variability in the determination of menstrual phase across all studies due to inconsistencies in defining and verifying the menstrual cycle phases as well as the timing of energy intake measurement throughout the menstrual cycle.¹⁵ It is important to note that only 3 studies used more than 1 method of menstrual phase verification, which is now considered to be insufficient in any current menstrual cycle research.¹⁵

It has been speculated that E₂ plays a role in energy intake through mechanisms related to appetite.¹³^,⁴⁸^,⁴⁹ E₂ receptors (ERs) are found as 2 subtypes: ER $α$ and ER $β$ ⁠,⁵⁰ where ER $α$ is the key E₂ receptor in the brain and is highly expressed on neurons within the arcuate nucleus,⁵¹ more specifically pro-opiomelanocortin (POMC), which triggers a signaling cascade to alter appetite and subsequent energy intake.⁵⁰ In rodents, as E₂ increases in the late FP towards its peak concentrations in the OP with no increase in P₄, POMC activity is increased and a resulting decrease in energy intake during the late FP/OP occurs.⁵² Previous work in female rats has demonstrated that removal of ER $α$ receptors results in increases in energy intake and the presence of active ER $α$ in the brainstem reduces energy intake,⁵³^,⁵⁴ further supporting that E₂ may bind to appetite-inhibiting neurons in the brain, which could explain the lower energy intake in the OP where a greater concentration of E₂ is present. It is important to note that, due to the lack of available energy intake data, the OP was excluded from this analysis, which highlights the need for future research including this short, but important, phase of the menstrual cycle.

Although the role of P₄ in appetite regulation is less understood,⁵⁵^,⁵⁶ mice supplemented with physiological doses of endogenous P₄ had increased 24-hour energy intake compared with those supplemented with E₂ only or a combination E₂ and P₄.⁵⁷ These results suggest that P₄ may have appetite-stimulating properties, subsequently increasing energy intake, or has inhibitory effects on appetite-inhibiting properties of E₂, possibly through antagonistic ER $α$ activity.¹⁶^,⁵⁵^,⁵⁸ It is important to note that, although rats are the animal model most often used to compare female hormonal fluctuations, the rat estrous cycle only lasts 4 to 5 days, has hormone fluctuations characterized by 4 distinct phases as opposed to 3, and female rats do not experience menstruation.⁴⁴ These key differences between female rats and humans are important to consider as they impact the mechanistic interaction of ovarian hormones and neuronal activity related to appetite.¹⁵ In the estrous cycle, rats experience a similar rise and peak of E₂ and P₄ in the proestrus phase of the estrous cycle, which is considered the first stage and only lasts approximately 14 hours,⁵⁹ reflecting similar hormone fluctuations in the LP in humans that lasts for approximately 14 days. However, rats do not experience a phase in which E₂ peaks independently of P₄,⁵⁹ highlighting the necessity for human research to understand the role of the ovarian hormones in appetite regulation and further supporting the need for the current study as well as future research on the role of ovarian hormones in appetite and energy intake research in females.

Energy intake was measured in all studies included for analysis via self-reported dietary records, which have previously been reported to be subject to bias,⁶⁰ despite remaining the most common method for measuring energy intake.⁵¹ Four of these studies attempted to improve validity of the food logs by weighing the foods consumed by each participant,¹⁹^,²²^,²⁵^,⁴³ 1 study provided a “menu” of food products from participants to choose from,²⁹ and only 1 study clearly defined that a registered dietitian had worked with participants to confirm energy intake.²³ These complexities highlight that the comparison of energy intake data validity between studies is difficult, especially in conjunction with inconsistent menstrual phase characterization. Recent methodological advances using imagery-based smartphone applications, particularly those monitored by a registered dietitian,⁶¹ may be helpful in ensuring energy intake data are consistent and reliable.^62–64 Eight of the studies included in this analysis asked participants to log their food on consecutive days,²⁰^,²²^,²⁵^,²⁶^,²⁹^,^41–43 while 3 studies did not state whether or not food logging was done on consecutive days or was completed sporadically.¹⁹^,²³^,³¹ From the 5 studies that measured energy intake daily, 3 studies found a greater energy intake in the LP compared with the FP²²^,⁴¹^,⁴² and 2 found no differences,²⁵^,²⁹ indicating that the period of time for which energy intake is measured may not be a predictor of energy intake across the menstrual cycle.¹⁵ Three of these studies that either found no differences between menstrual phases or greater energy intake in the FP compared with the LP also had significant individual variability in energy intake of up to approximately 945 calories per day¹⁰^,²⁹^,³¹ and 1 of these studies was underpowered.²⁹ In addition, 4 of these studies did not provide or report clearly the approximate day of testing phase for all phases.¹⁰^,²⁵^,²⁹^,³¹

While this systematic review and meta-analysis is the first to systematically synthesize the available data on energy intake across the menstrual cycle, there are several limitations to acknowledge. First, participants who completed at home dietary records may have over- or underestimated their energy intakes, which may have an influence on the mean differences between the FP and the LP.⁶⁵^,⁶⁶ Second, it is worth considering that energy intake may not be uniform across the week, as food consumption collected over the weekends compared with weekdays may differ²⁴ and this was not accounted for in many of the studies included in analysis. Third, it is important to consider that the follicular phase contains a hormonal profile where E₂ and P₄ are low for most of the phase; however, towards ovulation, E₂ begins to increase. Because of the lack of clarity within some studies on how exactly menstrual phase is determined and taking into consideration interindividual variability, it remains difficult to ensure that all participants maintained a low hormonal profile in the FP. Fourth, there is a possibility that this meta-analysis is unable to account for data that were collected but not published and we cannot completely exclude the possibility of publication bias. Finally, the current review includes literature that dates back as early as 1988, and due to the less detailed standard of menstrual phase determination in prior years, many technical flaws are present in these studies.¹⁵ With the recent emphasis on the inclusion of females in research studies, it is important that researchers account for hormonal patterns related to the menstrual cycle and apply effective methods for determining menstrual phase. Considering the small quantity of research pertaining to the inclusion criteria, no study was excluded based on date of publication.

CONCLUSION

In conclusion, this systematic review and meta-analysis determined that energy intake is greater in the LP compared with the FP, with a crude difference of approximately 168 kcal⋅d⁻¹. The overall effect size for differences in energy intake between phases moderately favors the LP having a higher energy intake, which may be attributed to the presence of elevated ovarian hormones during the LP and supports similar findings found in animal studies,^53–55 and is possibly explained by the need for females to maintain a neutral/positive energy balance in the LP in order to support a potential pregnancy.^45–47 There are methodological inconsistencies across the existing literature pertaining to both how menstrual phase is characterized and measurement of energy intake, suggesting that future work using robust methods to identify the menstrual cycle phase and best methods for the determination of energy intake is warranted. This study provides valuable insight on the effect of the menstrual cycle on energy intake, supports previous data using animal models, and further highlights the need for more rigorous research focused on female-specific physiology in hopes of improving the understanding of the menstrual cycle, specifically the fluctuations of ovarian hormones, and potential effects on energy intake.

Author contributions

J.A.L.T., S.F.M., D.P.D.B., and T.J.H. designed the research study. J.A.L.T., J.S.K., D.P.D.B., and S.F.M. conducted the research (search strategy, literature search, data analysis, risk of bias analysis, and data interpretation). J.A.L.T. and T.J.H. wrote the manuscript. All authors had responsibility for the final content.

Supplementary Material

Supplementary Material is available at Nutrition Reviews online.

Funding

No sources of funding were required to complete this project.

Conflicts of interest

None declared.

Data availability

All data analyzed in this review as well as searching protocol can be made available upon request. This review was not registered.

REFERENCES

Donnelly

Smith

BK.

Is exercise effective for weight loss with ad libitum diet? Energy balance, compensation, and gender differences

Exerc Sport Sci Rev

2005

;

(

169

174

10.1097/00003677-200510000-00004

Thomas

Bouchard

Church

, et al.

Why do individuals not lose more weight from an exercise intervention at a defined dose? An energy balance analysis

Obes Rev

2012

;

(

835

847

10.1111/j.1467-789X.2012.01012.x

Westerterp

Meijer

Janssen

Saris

Ten Hoor

Long-term effect of physical activity on energy balance and body composition

Br J Nutr

1992

;

(

Donnelly

Hill

Jacobsen

, et al.

Effects of a 16-month randomized controlled exercise trial on body weight and composition in young, overweight men and women: the Midwest Exercise Trial

Arch Intern Med

2003

;

163

(

1343

1350

10.1001/archinte.163.11.1343

Williams

Wood

Collins

Callister

Effectiveness of weight loss interventions—is there a difference between men and women: a systematic review

Obes Rev

2015

;

(

171

186

Gorin

Raynor

Fava

, et al.

Randomized controlled trial of a comprehensive home environment-focused weight-loss program for adults

Health Psychol

2013

;

(

128

137

Irving

Weltman

Patrie

, et al.

Effects of exercise training intensity on nocturnal growth hormone secretion in obese adults with the metabolic syndrome

J Clin Endocrinol Metab

2009

;

(

1979

1986

Tate

Jeffery

Sherwood

Wing

RR.

Long-term weight losses associated with prescription of higher physical activity goals. Are higher levels of physical activity protective against weight regain?

Am J Clin Nutr

2007

;

(

954

959

10.1093/ajcn/85.4.954

Wood

Stefanick

Williams

Haskell

WL.

The effects on plasma lipoproteins of a prudent weight-reducing diet, with or without exercise, in overweight men and women

N Engl J Med

1991

;

325

(

461

466

10.1056/nejm199108153250703

Elliott

Leow

Henry

CJ.

The influence of the menstrual cycle on energy balance and taste preference in Asian Chinese women

Eur J Nutr

2015

;

(

1323

1332

10.1007/s00394-014-0812-y

Rogan

Black

KE.

Dietary energy intake across the menstrual cycle: a narrative review

Nutr Rev

2022

;

(

869

886

10.1093/nutrit/nuac094

10.1016/j.appet.2022.106425

Geiker

Ritz

Pedersen

Larsen

Hill

Astrup

A weight-loss program adapted to the menstrual cycle increases weight loss in healthy, overweight, premenopausal women: a 6-mo randomized controlled trial

Am J Clin Nutr

2016

;

104

(

10.3945/ajcn.115.126565

Moniz

McCarthy

Broad

Medeiros

Hazell

TJ.

The exercise-induced suppression of acylated ghrelin is blunted in the luteal phase of the menstrual cycle compared to the follicular phase following vigorous-intensity exercise

Appetite

2023

;

182

106425

Buffenstein

Poppitt

McDevitt

Prentice

AM.

Food intake and the menstrual cycle: a retrospective analysis, with implications for appetite research

Physiol Behav

1995

;

(

1067

1077

10.1016/0031-9384(95)02003-9

Elliott-Sale

Minahan

de Jonge

, et al.

Methodological considerations for studies in sport and exercise science with women as participants: a working guide for standards of practice for research on women

Sports Med

2021

;

(

843

861

10.1007/s40279-021-01435-8

Hirschberg

AL.

Sex hormones, appetite and eating behaviour in women

Maturitas

2012

;

(

248

256

10.1016/j.maturitas.2011.12.016

Asarian

Geary

Sex differences in the physiology of eating

Am J Physiol Regul Integr Comp Physiol

2013

;

305

(

R1215

10.1152/ajpregu.00446.2012

Ajayi

Akhigbe

RE.

Staging of the estrous cycle and induction of estrus in experimental rodents: an update

Fertil Res Pract

2020

;

(

10.1186/s40738-020-00074-3

Cheikh Ismail

Al-Hourani

Lightowler

Aldhaheri

Henry

CJ.

Energy and nutrient intakes during different phases of the menstrual cycle in females in the United Arab Emirates

Ann Nutr Metab

2009

;

(

124

128

Chappell

Hackney

AC.

Associations between menstrual cycle phase, physical activity level and dietary macronutrient intake

Biol Sport

1997

;

(

251

258

PubMed

OpenURL Placeholder Text

10.1016/j.eclnm.2009.08.002

Gil

YRC

Fagundes

RLM

Santos

Calvo

MCM

Bernardine

JD.

Relation of menstrual cycle and alimentary consumption of women

E-SPEN, the European e-Journal of Clinical Nutrition and Metabolism

2009

;

(

E257

E260

10.1016/j.ando.2016.07.001

Gong

Garrel

Calloway

DH.

Menstrual cycle and voluntary food intake

Am J Clin Nutr

1989

;

(

252

258

10.1093/ajcn/49.2.252

Kammoun

Ben Saâda

Sifaou

, et al.

Change in women's eating habits during the menstrual cycle

Ann Endocrinol (Paris)

2017

;

(

Tsang

Lui

SS.

Menstrual cycle and voluntary food intake in young Chinese women

Appetite

1999

;

(

109

118

10.1006/appe.1999.0235

Lyons

Truswell

Mira

Vizzard

Abraham

SF.

Reduction of food intake in the ovulatory phase of the menstrual cycle

Am J Clin Nutr

1989

;

(

1164

1168

10.1093/ajcn/49.6.1164

Martini

Lampe

Slavin

Kurzer

MS.

Effect of the menstrual cycle on energy and nutrient intake

Am J Clin Nutr

1994

;

(

895

899

10.1093/ajcn/60.6.895

Ihalainen

Löfberg

Kotkajuuri

Kyröläinen

Hackney

Taipale-Mikkonen

RS.

Influence of menstrual cycle or hormonal contraceptive phase on energy intake and metabolic hormones—a pilot study

Endocrines

2021

;

(

10.3390/endocrines2020008

Bryant

Truesdale

Dye

Modest changes in dietary intake across the menstrual cycle: Implications for food intake research

Br J Nutr

2006

;

(

888

894

Fong

Kretsch

MJ.

Changes in dietary intake, urinary nitrogen, and urinary volume across the menstrual cycle

Am J Clin Nutr

1993

;

(

Gorczyca

Sjaarda

Mitchell

, et al.

Changes in macronutrient, micronutrient, and food group intakes throughout the menstrual cycle in healthy, premenopausal women

Eur J Nutr

2016

;

(

1181

1188

10.1007/s00394-015-0931-0

Souza

Martins

Cordeiro

Rodrigues

Rafacho

BPM

Bomfim

RA.

Do food intake and food cravings change during the menstrual cycle of young women?

Rev Bras Ginecol Obstet

2018

;

(

686

692

10.1055/s-0038-1675831

Cross

Marley

Miles

Willson

Changes in nutrient intake during the menstrual cycle of overweight women with premenstrual syndrome

Br J Nutr

2001

;

(

475

482

Dye

Blundell

JE.

Menstrual cycle and appetite control: implications for weight regulation

Hum Reprod

1997

;

(

1142

1151

10.1093/humrep/12.6.1142

Janse

DEJX

Thompson

Han

Methodological recommendations for menstrual cycle research in sports and exercise

Med Sci Sports Exerc

2019

;

(

2610

2617

10.1249/mss.0000000000002073

Hutton

Salanti

Caldwell

, et al.

The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations

Ann Intern Med

2015

;

162

(

777

784

Gold

EB.

The timing of the age at which natural menopause occurs

Obstet Gynecol Clin North Am

2011

;

(

425

440

10.1016/j.ogc.2011.05.002

Benton

Hutchins

Dawes

JJ.

Effect of menstrual cycle on resting metabolism: a systematic review and meta-analysis

PLoS One

2020

;

(

e0236025

10.1371/journal.pone.0236025

Altman

Bland

JM.

Standard deviations and standard errors

BMJ

2005

;

331

(

7521

903

10.1136/bmj.331.7521.903

Harrer

Cuijpers

Furukawa

Ebert

DD.

Doing Meta-analysis With R: A Hands-On Guide

. 1st ed.

Chapman & Hall/CRC Press

;

2021

Frampton

Edinburgh

Ogden

Gonzalez

Chambers

ES.

The acute effect of fasted exercise on energy intake, energy expenditure, subjective hunger and gastrointestinal hormone release compared to fed exercise in healthy individuals: a systematic review and network meta-analysis

Int J Obes (Lond)

2022

;

(

255

268

10.1038/s41366-021-00993-1

Johnson

Corrigan

Lemmon

Bergeron

Crusco

AH.

Energy regulation over the menstrual cycle

Physiol Behav

1994

;

(

523

527

10.1016/0031-9384(94)90296-8

Eck

Bennett

Egan

, et al.

Differences in macronutrient selections in users and nonusers of an oral contraceptive

Am J Clin Nutr

1997

;

(

419

424

10.1093/ajcn/65.2.419

Barr

Janelle

Prior

JC.

Energy intakes are higher during the luteal phase of ovulatory menstrual cycles

Am J Clin Nutr

1995

;

(

Asarian

Geary

Modulation of appetite by gonadal steroid hormones

Philos Trans R Soc Lond B Biol Sci

2006

;

361

(

1471

1251

1263

10.1098/rstb.2006.1860

Chen

Zhou

, et al.

The effects of weight loss-related amenorrhea on women's health and the therapeutic approaches: a narrative review

Ann Transl Med

2023

;

(

132

Oosthuyse

Bosch

AN.

The effect of the menstrual cycle on exercise metabolism: implications for exercise performance in eumenorrhoeic women

Sports Med

2010

;

(

207

227

10.2165/11317090-000000000-00000

Ryterska

Kordek

Załęska

Has menstruation disappeared? Functional hypothalamic amenorrhea—what is this story about?

Nutrients

2021

;

(

):2827.

10.3390/nu13082827

OpenURL Placeholder Text

10.1016/j.physbeh.2009.06.010

Butera

PC.

Estradiol and the control of food intake

Physiol Behav

2010

;

(

175

180

Dalvit

SP.

The effect of the menstrual cycle on patterns of food intake

Am J Clin Nutr

1981

;

(

1811

1815

10.1093/ajcn/34.9.1811

Mauvais-Jarvis

Clegg

Hevener

AL.

The role of estrogens in control of energy balance and glucose homeostasis

Endocr Rev

2013

;

(

309

338

Archundia Herrera

Chan

CB.

Narrative review of new methods for assessing food and energy intake

Nutrients

2018

;

(

):1064.

10.3390/nu10081064

OpenURL Placeholder Text

10.1016/j.cmet.2011.08.009

Wise

Scarbrough

Weiland

Larson

GH.

Diurnal pattern of proopiomelanocortin gene expression in the arcuate nucleus of proestrous, ovariectomized, and steroid-treated rats: a possible role in cyclic luteinizing hormone secretion

Mol Endocrinol

1990

;

(

886

892

Nedungadi

Zhu

, et al.

Distinct hypothalamic neurons mediate estrogenic effects on energy homeostasis and reproduction

Cell Metab

2011

;

(

453

465

Priego

Sánchez

Picó

Palou

Sex-associated differences in the leptin and ghrelin systems related with the induction of hyperphagia under high-fat diet exposure in rats

Horm Behav

2009

;

(

10.1016/j.yhbeh.2008.07.010

Klump

Keel

Racine

, et al.

The interactive effects of estrogen and progesterone on changes in emotional eating across the menstrual cycle

J Abnorm Psychol

2013

;

122

(

131

137

Vigil

Meléndez

Petkovic

Del Río

JP.

The importance of estradiol for body weight regulation in women

Front Endocrinol (Lausanne)

2022

;

951186

10.3389/fendo.2022.951186

Geary

Corwin

RL.

Individual effects of estradiol and progesterone on food intake and body weight in ovariectomized binge rats

Physiol Behav

2011

;

104

(

687

693

10.1016/j.physbeh.2011.07.017

Wade

GN.

Some effects of ovarian hormones on food intake and body weight in female rats

J Comp Physiol Psychol

1975

;

(

183

193

Sato

Nasu

Tsuchitani

Comparative histopathology of the estrous or menstrual cycle in laboratory animals

J Toxicol Pathol

2016

;

(

155

162

10.1293/tox.2016-0021

Rossato

Fuchs

SC.

Handling random errors and biases in methods used for short-term dietary assessment

Rev Saude Publica

2014

;

(

845

850

10.1590/s0034-8910.2014048005154

Bouzo

Plourde

Beckenstein

Cohen

TR.

Evaluation of the diet tracking smartphone application Keenoa™: a qualitative analysis

Can J Diet Pract Res

2022

;

(

10.3148/cjdpr-2021-022

Ferrara

Kim

Lin

Hua

Seto

A focused review of smartphone diet-tracking apps: usability, functionality, coherence with behavior change theory, and comparative validity of nutrient intake and energy estimates

JMIR Mhealth Uhealth

2019

;

(

e9232

Gioia

Vlasac

Babazadeh

, et al.

Mobile apps for dietary and food timing assessment: evaluation for use in clinical research

JMIR Form Res

2023

;

e35858

Hebden

Cook

van der Ploeg

Allman-Farinelli

Development of smartphone applications for nutrition and physical activity behavior change

JMIR Res Protoc

2012

;

(

Cook

Pryer

Shetty

The problem of accuracy in dietary surveys. Analysis of the over 65 UK National Diet and Nutrition Survey

J Epidemiol Commun Health

2000

;

(

611

616

10.1136/jech.54.8.611