Skip to main content
NCBI home page
Advances in Nutrition logoLink to Advances in Nutrition
. 2019 Jan 17;10(2):262–279. doi: 10.1093/advances/nmy074

Increased Dairy Product Intake Modifies Plasma Glucose Concentrations and Glycated Hemoglobin: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Sarah O'Connor 1,2, Anne-Frédérique Turcotte 1,3, Claudia Gagnon 1,3,4, Iwona Rudkowska 1,2,✉
PMCID: PMC6416042  PMID: 30657868

ABSTRACT

Dairy product intake is inversely associated with the risk of type 2 diabetes (T2D) in numerous cohort studies; yet, the beneficial effects of increased dairy product intake on T2D risk factors such as fasting plasma glucose, fasting insulin, insulin resistance with the homeostasis model assessment, and glycated hemoglobin (HbA1c) remain inconclusive in clinical trials. The objective of this study was to systematically review clinical trials observing the effects of elevated compared with minimal intake of dairy products on T2D risk factors in subjects without diabetes. Five databases [Medline, EMBASE, Central, CINAHL, AMED (Allied and Complementary Medicine)] were searched to identify randomized controlled trials that used elevated quantities of dairy products from ruminant sources in comparison with a lower intake in control groups. The review outcomes were fasting blood glucose, fasting insulin, homeostasis model assessment of insulin resistance (HOMA-IR), and HbA1c. Risk of bias and quality of evidence according to Grading of Recommendations Assessment, Development, and Evaluation were addressed. From the 10,627 citations screened, 44 studies (3016 participants) were included, 38 of which were used in the meta-analyses. Fasting glucose was positively associated with elevated dairy intake [34 studies, n = 2678; mean difference (MD): 0.07 mmol/L; 95% CI: 0.01, 0.12 mmol/L; P = 0.01, I2 = 23%]. Fasting insulin (29 studies, n = 1902; MD: −2.97 pmol/L; 95% CI: −7.05, 1.10 pmol/L; P = 0.15, I2 = 21%) and HOMA-IR (13 studies, n = 840; standardized MD: −0.07; 95% CI: −0.26, 0.12; P = 0.49, I2 = 38%) were not associated with elevated dairy consumption. HbA1c was negatively associated with elevated dairy product intake in 4 studies (n = 512; MD: −0.09%; 95% CI: −0.09%, −0.03%; P = 0.005, I2 = 0%). Most studies had high risk of bias and the quality of evidence was very low or low. In conclusion, evidence suggests that elevated dairy product intake is associated with increased fasting plasma glucose concentrations together with reduced HbA1c in nondiabetic subjects. Hence, the clinical significance of these results remains uncertain. Additional well-designed, long-term studies are required.

Keywords: dairy products, milk, yogurt, cheese, prediabetes, glycemia, insulinemia, insulin resistance, glycated hemoglobin

Introduction

Diabetes is a chronic disease that affected 422 million people worldwide in 2014 (1). In 2012, ∼1.5 million deaths were attributed to diabetes and its complications (1). The accelerating rate of diabetes prevalence represents a significant global economic burden and is mainly due to type 2 diabetes (T2D), which represents the vast majority of all diabetes diagnoses (2, 3). T2D was previously diagnosed almost exclusively during adulthood and in the elderly; however, there is an increasing prevalence in childhood and adolescence (1, 4). T2D is characterized by an ineffective response of body cells to insulin along with an insufficient insulin secretion, which results in hyperglycemia (5, 6). The onset of T2D is usually preceded by insulin resistance and hyperinsulinemia in order to keep blood glucose concentrations within the normal range (5, 6). When insulin resistance is no longer manageable by elevated insulin secretion, fasting blood glucose increases until it reaches the pathologic concentrations of T2D (5). Current T2D diagnostic criteria include fasting plasma glucose concentrations ≥7.0 mmol/L, plasma glucose concentrations ≥11.1 mmol/L 2 h after ingestion of 75 g glucose, and glycated hemoglobin (HbA1c) ≥6.5% (5). Insulin resistance can be evaluated with the use of HOMA-IR, a commonly used index that has a good correlation with gold-standard measures (7). Normal cutoff values of the HOMA-IR index vary between 1.4 and 2 according to ethnic group and medical condition (8).

Dairy products are complex foods containing specific nutrients that could have beneficial effects on T2D risk factors. Among these nutrients, BCAAs contained in whey protein in dairy products are known to stimulate insulin secretion (9). Further, the dairy matrix contains significant amounts of calcium and vitamin D and a small amount of unique ruminant trans FAs, which have been associated with reduced insulin resistance and improved insulin sensitivity (10, 11). Increasing interest in dairy products for the prevention of T2D has emerged following the publication of systematic reviews and meta-analyses of observational studies suggesting a reduction in T2D incidence with dairy product intake ≤200–400 g/d, mostly attributable to yogurt and low-fat dairy products (12–15). From this observational evidence, current guidelines for the prevention of T2D include moderate low-fat dairy product intake as part of a healthy diet, including milk, yogurt, and cheese (16, 17). In contrast to observational evidence, systematic reviews of randomized trials failed to observe clinical benefits of dairy product intake on T2D risk factors, including fasting glucose and the HOMA-IR index (18, 19). In addition, the clinical role of elevated dairy product intake on HbA1c remains unaddressed in previous systematic reviews (18, 19). Thus, the main objective of this current study is to systematically review randomized controlled trials assessing different quantities of dairy products and to observe if an elevated dairy product intake can affect T2D risk factors in comparison with minimal intake in subjects without diabetes.

Methods

The systematic review was conducted in accordance with the Cochrane Handbook for Systematic Reviews of Interventions (20). The review protocol was registered with the Prospective Register of Systematic Reviews (PROSPERO) on 4 December 2017 (no. CRD42017080137) (21). The Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) structure was used for the review presentation (22)

Inclusion and exclusion criteria

Randomized controlled trials of ≥7 d were included. Participants were humans of any gender or ethnicity, excluding pregnant or lactating women and subjects suffering from any type of diabetes or using diabetes medication. Children >5 y old, representing ≥80% of the subjects in the studies, were included to take into account early-onset T2D. Studies were included when the exposure group was administered a higher quantity of dairy products than the control group; therefore, both groups had either elevated or minimal dairy product intake. The groups with the highest daily intake of dairy products were considered the exposure groups. All types of dairy products from ruminant animals taken orally were included, with the exclusion of infant formulas, human milk, milk colostrum, dairy substitutes, and dairy nutrients extracted from the dairy matrix (butter, milk proteins, milk fat, etc.). The groups with the lowest intake of dairy products or assigned to receive nondairy foods or placebo known to have no influence on glucose metabolism were considered control groups (minimal intake). No threshold was imposed for the distinction between elevated and minimal intake. Trials with cointerventions were included only if the total intake of dairy products remained different between groups, and the cointerventions were the same in both elevated dairy and minimal dairy groups. The primary outcomes were fasting blood glucose concentrations, fasting insulin concentrations, HbA1c, and insulin resistance measured by the HOMA-IR (7). Only end-point measures were considered. Regarding HbA1c, only studies of ≥12 wk were eligible due to the 3-mo turnover rate of hemoglobin (5).

Search methods and study selection

Electronic searches

Only peer-reviewed publications from inception until 25 October 2017 were considered. No restriction on language was imposed. The following databases were searched on 25 October 2017: Central, Medline, EMBASE, Allied and Complementary Medicine (AMED), and CINAHL. The search strategy is available online (Supplemental Material 1). Indexed and free terms were used to define the population and exposure. An elevated sensitivity and precision maximizing filter for clinical trials was used for Medline, EMBASE, and CINAHL (20, 23, 24). The reference lists of included studies were searched for other relevant reports. Protocols were retrieved from online protocol databases or by contacting the authors, when available.

Study selection

Results from the search strategies were merged and duplicates were removed with EndNote version X8.1 (Clarivate Analytics). Two reviewers (SO and A-FT) independently screened titles and abstracts in duplicate. Study eligibility was determined independently in duplicate (SO and A-FT) through the use of full-text reports. Publications from the same authors and locations were clustered and compared to identify complementary articles from the same studies. In case of disagreement, a third reviewer was invited in the discussion.

Data extraction

An adapted data collection form was pilot-tested prior to data extraction. The collection form was inspired by the Data Collection Form for RCTs from Cochrane and the Cochrane Handbook (20, 25). Data were extracted independently by 2 reviewers (SO and A-FT) in duplicate. Extracted characteristics included country, study design, number of participants, exclusion and withdrawal information, age, sex, BMI, comorbidities, exposure and comparator details (type and doses of dairy products, length of exposure, cointerventions), administration of dairy products (free-living condition or through a controlled diet), and funding sources. The number of subjects per group, outcome measures, and deviation measures were extracted for meta-analyses. In case of duplicate reporting, the reports with the largest number of participants were used. In case of missing information, corresponding authors were contacted twice. Internal validity for each study was confirmed based on the criteria of the risk-of-bias tool from the Cochrane Handbook (20). The risk of bias for each domain was rated as “low,” “high,” or “unclear.”

Data analysis

Studies with parallel or cluster designs were included in meta-analyses. For crossover designs, only the data from the first intervention period were used in order to reduce potential bias from the carry-over effect (20). Corresponding authors were contacted when the data from the first intervention period were not available in reports. Fasting glucose (millimoles per liter), fasting insulin (picomoles per liter), and HbA1c (%) were analyzed as continuous outcomes and presented as mean differences (MDs) with 95% CIs through the use of the inverse-variance method. Standardized MD (SMD) was used for insulin resistance because a correction of scale measurement was necessary (20). When only the difference from baseline was available, the final measurements were estimated from baseline values and the variation measures from baseline were used (20). Statistical heterogeneity was assessed through the use of the I2 statistic (26); I2 >50% was considered as substantial heterogeneity (27). Average dairy product intake was calculated from the servings proposed in studies eligible for the meta-analyses or were estimated according to the serving sizes of the Canada's Food Guide for Healthy Eating 2007 (28). Subgroup analysis was performed to investigate potential sources of inconsistency: obese or nonobese [with BMI (in kg/m2) ≥30 as the threshold for obesity], presence or absence of energy restriction, trial length (<24 or ≥24 wk), presence or absence of metabolic disorders, funding source (industry or independent), and risk of bias (low/unclear or elevated). Post hoc subgroups were performed for each continent and Northern/Southern Europe, liquid (milk, yogurt) compared with solid (cheese), low-fat (<2% fat content) compared with high-fat (≥2% of fat content), fermented (yogurt, cheese) compared with nonfermented (milk), free-living interventions compared with controlled diets, and glucose metabolism as a primary outcome compared with not a primary outcome. All analyses were performed with the use of a random-effects model. A P value <0.05 was considered statistically significant. Analyses were realized with Review Manager software (29). When data extraction could not be performed, the studies were used for qualitative analyses.

Quality of evidence and publication bias

Publication bias was assessed by visual evaluation of funnel plots produced by Review Manager software (29). Quality of evidence for each outcome was assessed according to the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach (30)

Results

In total, 10,627 citations were identified and screened after the removal of duplicates (Figure 1). From the potential citations, 386 full-text reports were retrieved, 102 of which were complementary articles of the same studies. From the 284 studies assessed for eligibility, 44 were included in the systematic review (31–73). One report included 2 distinct eligible randomized controlled trials (71). Six studies were excluded from the meta-analyses and used for qualitative analysis only: 5 crossover trials without available data after the first period of intervention (39, 47, 55, 57, 61) and 1 trial for which data could not be extracted (IQRs as a variation measure) (36). Thirty-eight studies were included in the meta-analyses (31–35, 37, 38, 40–46, 48–54, 56, 58–60, 62–73). Data from 5 crossover studies were retrieved through corresponding authors (38, 40, 42, 52, 53). In 6 studies with multiple exposures, groups with similar servings of dairy products were pooled (38, 41, 48, 50, 52, 54). Final values were estimated with change from baseline measures in 7 studies (33, 51, 54, 58, 62, 65, 73). The difference from baseline and variation measures of final measurement were used in 1 study since no estimation was possible (56). Data and variation measures were estimated from figures in 5 studies (33, 37, 70, 71). In 1 study, the number of subjects was presented as ranges; therefore, the lowest number of participants was used (69).

FIGURE 1.

FIGURE 1

Open in a new tab

Flow diagram of identified citations and included studies.

Study characteristics

Study characteristics are presented in Table 1. All studies were published in English. Sample sizes varied from 10 to 205 subjects, giving a total of 3016 participants with a mean ± SD age of 41.4 ± 1.4 y and a BMI of 39.4 ± 0.8. The pooled average intake of dairy products was 3.0 servings/d in the exposure group (elevated dairy product intake) and the mean intake was 0.5 servings/d in the minimal intake group. Twenty-three studies were from North America (34, 38, 39, 42, 44, 45, 49, 50, 52, 53, 55, 57, 59, 60, 62, 63, 65, 68, 70–73), 8 from Europe (33, 35, 46, 47, 54, 61, 67, 69), 4 from Oceania (37, 40, 43, 48), 5 from South America (31, 32, 36, 58, 66), and 4 from Asia (41, 51, 53, 56). The exposure lengths varied from 1 to 48 wk and subjects from 10 studies were exposed to dairy products for ≥24 wk (35, 45, 54, 57, 59, 64, 65, 69–72). Main cointerventions were probiotics (36, 46, 52, 56, 67) and energy restriction programs (35, 37, 39, 43, 45, 49, 50, 55, 58–60, 65, 66, 68, 70–73). Three studies assessed glucose metabolism parameters as primary outcomes (33, 46, 67), whereas 14 studies had other related metabolic parameters as primary outcomes (34, 37, 38, 41–43, 48, 51, 57, 59, 63, 64, 66, 69). Twenty-four protocols were retrieved through online databases. The risk-of-bias assessment results for included studies is available online (Supplemental Figure 1). The overall risk of bias was considered “high” for all except 12 studies, which had an unclear risk of bias (33, 35, 38, 42, 49–51, 53, 61, 64, 67).

TABLE 1.

Characteristics of included studies on dairy product intake and type 2 diabetes–related parameters1

Study (reference) Country Age at baseline, y Sex, n male/total BMI, kg/m2 Comorbidity Length, wk Context Dairy exposure (n of subjects) Control (n of subjects) Energy restriction
Alemán-Mateo et al. (31) Mexico 76.0 ± 5.4 17/40 26.3 ± 3.8 Sarcopenia 12 Free-living 3 × 70 g ricotta cheese (n = 20) Habitual diet (n = 20) None
Alemán-Mateo et al. (32) Mexico 70.2 ± 7.0 50/100 27.1 ± 3.5 None 12 Free-living 3 × 70 g ricotta cheese (n = 50) Habitual diet (n = 50) None
Arnberg et al. (33) Denmark 13.2 ± 7.0 37/98 25 ± 1.2 None 12 Free-living 1 L skimmed milk (n = 48) 1 L water (n = 50) None
Barr et al. (34) Canada 65.2 ± 6.6 71/200 26.1 ± 3.5 None 12 Free-living 3 × 237 mL milk (n = 98) <1.5 serving of dairy products (n = 102) None
Benatar et al. (48) New Zealand 46.7 ± 11.7 65/180 24.5 ± 4.0 None for majority 4 Free living Increased intake (3–4 extra servings) of dairy products (n = 60) No change from normal diet (n = 60) None
Decreased dairy intake (minimal intake) (n = 60)
Bendtsen et al. (35) Denmark 43.5 ± 7.1 11/80 31.1 ± 1.8 None 24 Free-living 1500 mg calcium/d, 1200 mg from dairy products (4–5 servings/d) (n = 22) 600 mg calcium maximum (0–1 servings/d) (n = 30) Yes
Bernini et al. (36) Brazil 18–60 (range) Males and females 33.3 (median) Metabolic syndrome 6.5 Free-living 80 mL fermented milk with probiotics (n = 26) No fermented milk (n = 25) None
Bowen et al. (37) Australia 47.7 ± 10.9 20/50 33.4 ± 11.9 None 16 Free-living Elevated dairy protein diet (500 mL milk, 30 g cheddar cheese, 200 g yogurt) (n = 25) Mixed-protein diet (125 mL milk, wholemeal bread, fruits, meat, almonds, legumes, ham) (n = 25) Yes
Brassard et al. (38) Canada 38.7 ± 13.5 49/922 31.0 ± 6.0 Low HDL cholesterol 4 Controlled Saturated fat diet from cheese (n = 64) Saturated fat from butter (n = 64) None
Elevated carbohydrate diet (n = 64)
Monounsaturated fat diet (n = 64)
Polyunsaturated fat diet (n = 64)
Buchowski et al. (39) USA 21–50 (range) 10/34 35.8 ± 4.3 None 6 Controlled Elevated calcium from dairy products (n = 34) Elevated calcium from nondairy food products (n = 34) Yes
Crichton et al. (40) Australia 49.0 ± 14.0 10/36 32 ± 6 None 12 Free-living 4 servings of low-fat dairy products (n = 36) 1 serving of low-fat dairy products (n = 36) None
Dönmez et al. (41) Kyrgyzstan 33.6 ± 3.17 18/18 Not reported None 2 Not reported 350 mL koumiss (n = 6) Physical activity (n = 6) None
350 mL koumiss + physical activity (n = 6)
Drouin-Chartier et al. (42) Canada 57.0 ± 5.0 0/27 31.9 ± 3.5 63% with the metabolic syndrome 6 Controlled 3.2 servings of milk per 2000 calories (n = 27) No milk or other dairy (n = 27) None
Farnsworth et al. (43) Australia 50.4 ± 2.9 14/57 34.1 ± 3.7 Hyperinsulinemia 16 60% controlled 250 mL milk, 30 g skimmed milk powder, 60 g cheese, 200 g yogurt + increased poultry (n = 28) 250 mL milk + increased fruits, wholemeal bread, rice, and biscuits (n = 29) Yes
Ghadirian et al. (44) Canada ≥50 0/158 Not reported None 4 Controlled Dairy food diet (∼30 g dairy protein) (n = 81) Dairy-free diet (n = 77) None
Gilbert et al. (45) Canada 41.0 ± 7.3 1/15 33.1 ± 3.1 None 24 Free-living 1 serving of milk enriched in calcium (n = 13) 1 serving of nonenriched calcium-free rice beverage (n = 12) Yes
Hulston et al. (46) UK 24.5 ± 5.8 14/15 23.9 ± 2.8 None 4 Free-living (3 wk) controlled (1 wk) 2 × 65 mL fermented milk (n = 8) No milk (n = 9) None
Jacobsen et al. (47) Denmark 24.2 ± 2.0 2/10 26.5 ± 2 None 1 Controlled Elevated calcium, elevated protein from low-fat dairy (n = 10) Low-calcium, low protein (n = 10) None
Jones et al. (49) Canada 51.1 ± 2.2 14/38 32.3 ± 3.7 Metabolic syndrome 12 Free-living 3–4 servings of milk or yogurt + calcium supplement (n = 20) 1 serving of milk or yogurt (n = 18) Yes
Josse et al. (50) Canada 28.0 ± 5.5 0/90 31.6 ± 3.3 None 12 Free-living 6–7 servings of dairy products (n = 30) 0–1 serving of dairy products (n = 30) Yes
3–4 servings of dairy products (n = 30)
Lee et al. (51) South Korea 50.0 ± 8.8 29/58 27.8 ± 2.9 Metabolic syndrome 6 Free-living 400 mL low-fat milk (n = 28) <200 mL/d, max 3×/wk (n = 30) None
Lee et al. (52) USA 28.0 ± 6.6 11/30 24.2 ± 2.7 None 4 Free-living 240 g yogurt smoothie (n = 30) Probiotic capsules (n = 30) None
240 g yogurt + probiotic added before fermentation (n = 30)
40 g yogurt + probiotic added after fermentation (n = 30)
Machin et al. (53) USA 53.2 ± 2.0 22/49 30.5 ± 7.0 Elevated blood pressure 4 Free-living 4 servings of dairy products (n = 49) 4 servings of fruits (n = 49) None
Maersk et al. (54) Denmark 38.8 ± 7.8 17/47 32.1 ± 12.5 None 24 Free-living 1 L semiskimmed milk (n = 12) 1 L water (n = 13) None
1 L cola (n = 10)
1 L diet cola (n = 12)
Melanson et al. (55) USA 33.5 ± 8.0 10/19 27.5 ± 3.5 None 1 Controlled 3–4 servings of dairy products, energy balance (n = 19) 1 serving of dairy products, energy balance (n = 19) Yes
3–4 servings of dairy products, energy restriction (n = 19) 1 serving of dairy products, energy restriction (n = 19)
Pu et al. (56) China 58.5 ± 8.3 59/205 25.8 ± 3.1 None 12 Free-living 3 × 100 mL fermented yogurt (n = 103) No yogurt (n = 102) None
Rideout et al. (57) Canada 53.0 ± 12.3 5/23 31.9 ± 3.0 None 24 Free-living 4 servings of dairy products (milk and yogurt) (n = 23) ≤2 servings of dairy products/d (n = 23) None
Rosado et al. (58) Mexico 34.5 ± 5.8 0/101 35.0 ± 4.3 None 16 Free-living 3 × 250 mL milk (n = 33) No milk (n = 31) Yes
3 × 250 mL milk + micronutrients (n = 37)
Shlisky et al. (59) USA 33.7 ± 7.3 0/80 29.2 ± 3.1 None 24 Free-living 2 × 177 mL nonfat yogurt + 3 servings of nonfat dairy products (n = 43) 3 servings of dairy products (n = 37) Yes
Smilowitz et al. (60) USA 24.8 ± 5.0 4/61 28.7 ± 2.8 None 12 Free-living 3 servings of dairy products (n = 22) Calcium supplement (n = 16) Yes
0–1 serving of dairy products (n = 23)
Soerensen et al. (61) Denmark 27.7 ± 4.8 15/15 23.1 ± 2.3 None 2 Controlled 670 mL semiskimmed milk (n = 15) Control diet (no dairy products except butter) (n = 15) None
120 g semihard cheese (n = 15)
Stancliffe et al. (62) USA 37.0 ± 9.9 19/40 30.7 ± 4.1 Metabolic syndrome 12 Free-living 3 servings of dairy products (n = 20) <0.5 serving of dairy products (n = 20) None
St-Onge et al. (63) USA 9.4 ± 5.3 9/45 27.4 ± 4.3 None 16 Free-living 4 × 236 mL skim milk (n = 21) 1 × 236 mL skim milk (n = 24) None
Tanaka et al. (64) Japan 41.7 ± 7.3 213/213 27.0 ± 3.4 Metabolic syndrome 24 Free-living 400 g milk or yogurt (n = 106) No dairy (n = 107) None
Thompson et al. (65) USA 41.4 ± 8.9 13/72 34.8 ± 3.1 None 48 Free-living 4 servings of dairy products (2 servings of milk) (n = 22) 2 servings of dairy products (n = 26) Yes
4 servings of dairy products (2 servings of milk) + fiber (n = 24)
Torres et al. (66) Brazil 40.4 ± 10.0 5/39 32.2 ± 1.5 None 16 Free-living Elevated-calcium diet + 60 g nonfat powdered milk (n = 19) Low-calcium diet (n = 20) Yes
Tripolt et al. (67) Austria 53 ± 7.1 18/28 33.2 ± 3.2 Metabolic syndrome 12 Free-living Probiotic milk drink, 3 × 65 mL/d (n = 13) Usual diet (n = 15) Yes
Van Loan et al. (68) USA 32.5 ± 9.5 0/71 32.5 ± 2.7 None 12 Controlled 3–4 servings of dairy products (n = 35) <1 serving of dairy products (n = 36) Yes
Wennersberg et al. (69) Norway 54.0 ± 7.8 37/113 30.1 ± 3.5 ≥2 components of the metabolic syndrome 24 Free-living 3–5 servings of dairy products (n = 57) ≤2 servings of dairy products (n = 56) None
Zemel et al. (70) USA 49.0 ± 6.0 5/32 34.9 ± 4.3 None 24 Not Reported 3 servings of dairy products (n = 11) Calcium supplement (n = 11) Yes
0–1 serving of dairy products (n = 10)
Zemel et al. Phase 1 (71) USA 41.9 ± 10.9 11/34 34.5 ± 3.1 None 24 Not reported 3 servings/d of dairy products (n = 17) 0–1 servings/d of dairy products (n = 17) None
Zemel et al. Phase 2 (71) USA 41.7 ± 10.7 4/29 35.5 ± 3.0 None 24 Not reported 3 servings of dairy products (n = 17) 0–1 serving of dairy products (n = 12) Yes
Zemel et al. (72) USA 40.3 ± 7.0 Males and females 34.5 ± 3.2 None 36 Free-living >3 servings of dairy products (n = 119) <1 serving of dairy products (n = 125) Yes
Zemel et al. (73) USA 25.7 ± 4.9 22/42 29.3 ± 2.7 None 12 Free-living 3 servings of dairy products (n = 23) Calcium supplement (n = 19) Yes
0–1 servings of dairy products (n = 26)
Open in a new tab

1Age and BMI are means ± SDs unless otherwise indicated.

2Number of subjects randomly assigned.

Fasting blood glucose

A summary of findings for all T2D risk factors is presented in Table 2. Elevated dairy product intake (3.1 servings/d, on average) was associated with increased fasting glucose compared with a minimal intake (0.5 serving/d, on average) (34 studies, n = 2678; MD: 0.07 mmol/L; 95% CI: 0.01, 0.12 mmol/L; P = 0.01, I2 = 23%)(Figure 2) (31–35, 37, 40, 41, 43–46, 48–54, 56, 58–60, 62–69, 72). Subgroup analyses revealed a positive association between fasting glucose concentrations and elevated dairy product intake in studies with low/unclear risk of bias (P < 0.0001) and with short-term exposure (<24 wk) (P = 0.02) (Table 3). Post hoc subgroup analyses suggest that studies from Asia/Middle East countries (P = 0.01) and Oceania (P = 0.002) were associated with increased fasting glucose concentrations with elevated dairy intake, along with liquid dairy products (P = 0.007), controlled diets during trials (P = 0.0002), and studies in which glucose metabolism was not the primary outcome (P = 0.03) (Table 3). The overall risk of bias was considered “high.” From the studies excluded from the meta-analysis, none of the 5 studies observed differences between groups (36, 39, 55, 57, 61).

TABLE 2.

Quality of evidence based on the GRADE approach and summary of findings1

GRADE approach Summary of findings
Outcomes Limitations of detailed design (risk of bias) Inconsistency (heterogeneity) Indirectness Imprecision (number of events, SD) Publication bias Decision Effect estimate (95% CI) Number of participants (studies) Quality of evidence
Fasting glucose Very serious limitations Consistent No major indirectness detected No major imprecision detected Unusual funnel plot Very low quality MD: 0.07 mmol/L (0.01, 0.12 mmol/L) 2678 (34) Very low quality
Fasting insulin Very serious limitations Consistent No major indirectness detected Some imprecision Unusual funnel plot Very low quality MD: −2.97 pmol/L (−7.05, 1.10 pmol/L) 1902 (29) Very low quality
HOMA-IR Very serious limitations Consistent No major indirectness detected No major imprecision detected Undetected Low quality SMD: −0.07 (−0.26, 0.12) 840 (13) Low quality
HbA1c Serious limitations Consistent Indirectness for participants No major imprecision detected Undetectable Low quality MD: −0.07% (−0.13%, −0.01%) 615 (6) Low quality
Open in a new tab

1The GRADE approach for quality of evidence: outcomes from randomized controlled trials start as high-quality evidence. Five factors may downrate the quality of evidence: limitation of detailed design (overall risk of bias of studies addressing the outcome), inconsistency (unexplained statistical heterogeneity unsolved by subgroup analyses), indirectness of evidence (indirect population, exposure and comparator, and outcomes), imprecision (wide CIs), probability of publication bias (based on funnel plots). GRADE, Grading of Recommendations Assessment, Development, and Evaluation; HbA1c, glycated hemoglobin; MD, mean difference; SMD, standardized mean difference.

FIGURE 2.

FIGURE 2

Open in a new tab

Forest plot of the meta-analysis on the effect of increased dairy product intake on fasting blood glucose concentrations. Values are weighted mean differences and corresponding 95% CIs in fasting glucose concentrations after elevated dairy product intake in comparison with a minimal intake in randomized controlled trials. Weighted mean differences were calculated with the inverse-variance method in a random-effects model. Each study is represented by gray squares, and the whiskers represent the 95% CIs. The black diamond represents the overall mean difference, the center of which corresponds to the pooled weight difference, and the extremities represent 95% CIs. Statistical heterogeneity is represented by the I2 statistic.

TABLE 3.

Subgroup analyses of included studies on dairy product intake and T2D-related parameters1

Outcome and subgroups Number of studies Number of participants (elevated dairy/control) Effect estimate (95% CI) I 2, % P
Fasting blood glucose Participant BMI (kg/m2)
  BMI ≥30 19 1202 (573/629) 0.05 (−0.02, 0.13) 10 0.15
  BMI <30 13 1300 (625/675) 0.05 (−0.02, 0.12) 26 0.13
 Dairy exposure combined with energy restriction
  No energy restriction 20 1571 (726/845) 0.08 (0.00, 0.15) 29 0.05
  Energy restriction 13 902 (462/440) 0.05 (−0.03, 0.13) 19 0.18
 Duration of exposure
  <24 wk 26 1885 (904/981) 0.08 (0.01, 0.14) 14 0.02
  ≥24 wk 8 807 (387/406) 0.04 (−0.06, 0.14) 29 0.41
 Comorbid condition
  No metabolic disorder 24 1981 (973/1008) 0.05 (−0.00, 0.10) 0 0.05
  Metabolic disorders 11 841 (414/478) 0.09 (−0.04, 0.21) 46 0.17
 Funding
  Industry 26 232 (1072/1111) 0.06 (−0.00, 0.12) 35 0.06
  Independent 7 428 (184/244) 0.09 (−0.02, 0.20) 0 0.12
 Risk of bias
  Unclear/low risk 10 856 (381/475) 0.14 (0.07, 0.21) 0 <0.0001
  Elevated risk 24 1822 (910/912) 0.03 (−0.04, 0.10) 29 0.40
Post hoc subgroup analyses for fasting blood glucose Continent
  North America 16 1288 (630/658) 0.05 (−0.04, 0.13) 35 0.29
  South America 4 243 (122/121) −0.06 (−0.23, 0.10) 0 0.45
  Northern Europe 4 294 (136/158) −0.01 (−0.12, 0.011) 0 0.91
  Southern Europe 2 45 (21/24) 0.15 (−0.16, 0.47) 0 0.35
  Middle East/Asia 4 481 (245/236) 0.13 (0.03, 0.24) 15 0.01
  Oceania 4 327 (137/190) 0.20 (0.07, 0.32) 0 0.002
 Fat content
  Low fat 14 862 (428/434) 0.07 (0.00, 0.14) 0 0.05
  High fat 1 91 (13/78) 0.13 (−0.22, 0.48) — 0.47
 Fermentation
  Fermented 8 535 (243/292) 0.02 (−0.09, 0.12) 0 0.76
  Nonfermented 9 594 (288/306) 0.05 (−0.03, 0.13) 0 0.26
 Consistency of dairy products
  Solid 3 231 (83/148) −0.04 (−0.23, 0.14) 0 0.64
  Liquid 16 1136 (570/566) 0.08 (0.02, 0.14) 0 0.007
 Glucose metabolism as primary outcome
  Primary outcome 3 143 (69/74) 0.07 (−0.06, 0.20) 0 0.31
  Not a primary outcome 31 2535 (1222/1313) 0.06 (0.01, 0.12) 29 0.03
 Type of dietary intervention
  Controlled diet 5 405 (174/231) 0.22 (0.10, 0.33) 0 0.0002
  Free-living (noncontrolled diet) 28 2255 (1105/1150) 0.04 (−0.01, 0.09) 15 0.13
Fasting insulin Participant BMI
  BMI ≥30 19 1124 (534/590) −5.49 (−11.74, 0.75) 33 0.08
  BMI <30 10 778 (357/421) −0.43 (−5.20, 4.34) 0 0.86
 Dairy exposure combined with energy restriction
  No energy restriction 14 1174 (513/661) −2.71 (−6.78, 1.35) 0 0.19
  Energy restriction 15 932 (477/455) −0.69 (−4.96, 3.59) 0 0.75
 Duration of exposure
  <24 wk 19 1434 (662/772) 0.30 (−3.38, 3.98) 0 0.87
  ≥24 wk 10 672 (328/344) −5.4 (−10.31, −0.49) 0 0.03
 Comorbid conditions
  No metabolic disorder 22 1512 (727/785) −2.08 (−5.47, 1.32) 0 0.23
  Metabolic disorders 7 594 (263/331) −0.75 (−6.69, 5.18) 0 0.80
 Funding
  Industry 24 1577 (761/816) −3.29 (−8.17, 1.58) 31 0.19
  Independent 5 325 (130/195) −2.72 (−9.58, 4.15) 0 0.44
 Risk of bias
  Unclear/low risk 7 549 (228/321) −0.84 (−6.79, 5.11) 0 0.78
  Elevated risk 22 1353 (663/690) −4.44 (−9.57, 2.36) 30 0.09
Post hoc subgroup analyses for fasting insulin Continent
  North America 18 1206 (586/620) −4.32 (−11.40, 2.76) 44 0.23
  South America 2 79 (39/40) 2.84 (−8.70, 14.38) 0 0.63
  Northern Europe 4 289 (134/155) −8.56 (−17.50, 0.39) 0 0.06
  Southern Europe 2 45 (21/24) 4.41 (−24.03, 32.85) 0 0.76
  Middle East/Asia 0 — — — —
  Oceania 3 283 (111/172) −1.36 (−7.72, 5.00) 0 0.68
 Fat content
  Low fat 10 646 (317/329) −2.03 (−10.09, 6.03) 0 0.62
  High fat 1 91 (13/78) 22.62 (−30.78, 76.02) — 0.41
 Fermentation
  Fermented 5 205 (71/134) 3.28 (−7.36, 13.91) 0 0.55
  Nonfermented 7 471 (226/245) −6.28 (−16.21, 3.66) 0 0.22
 Consistency of dairy products
  Solid 2 131 (33/98) 5.09 (−6.78, 16.96) 0 0.40
  Liquid 11 583 (284/299) −5.08 (−14.14, 3.99) 0 0.27
 Glucose metabolism as primary outcome
  Primary outcome 3 143 (69/74) −0.83 (−19.13, 17.47) 0 0.93
  Not a primary outcome 26 1759 (822/937) −3.18 (−7.61, 1.26) 29 0.16
 Type of dietary intervention
  Controlled diet 4 247 (92/155) 9.76 (−6.89, 26.41) 0 0.25
  Free-living (noncontrolled diet) 22 1571 (754/817) −2.42 (−0.675, 1.73) 19 0.25
HOMA-IR Participant BMI
  BMI ≥30 8 452 (186/266) −0.09 (−0.41, 0.24) 60 0.60
  BMI <30 5 388 (165/223) −0.04 (−0.24, 0.17) 0 0.71
 Dairy exposure combined with energy restriction
  No energy restriction 9 647 (255/392) −0.16 (−0.40, 0.09) 47 0.21
  Energy restriction 4 193 (96/97) 0.15 (−0.14, 0.43) 0 0.31
 Duration of exposure
  <24 wk 2 145 (66/79) −0.33 (−0.66, 0.00) 0 0.05
  ≥24 wk 11 695 (285/410) −0.02 (−0.23, 0.20) 39 0.88
 Comorbid conditions
  No metabolic disorder 8 535 (231/304) −0.02 (−0.20, 0.15) 0 0.78
  Metabolic disorders 5 305 (120/185) −0.12 (−0.63, 0.40) 74 0.65
 Funding
  Industry 9 559 (241/318) −0.12 (−0.38, 0.14) 52 0.37
  Independent 4 281 (110/171) 0.00 (−0.24, 0.24) 0 1.00
 Risk of bias
  Unclear/low risk 5 431 (152/279) 0.06 (−0.15, 0.27) 0 0.57
  Elevated risk 8 409 (199/210) −0.20 (−0.49, 0.08) 48 0.16
Post hoc subgroup analyses for HOMA-IR Continent
  North America 6 314 (127/187) −0.06 (−0.49, 0.38) 68 0.80
  South America 2 79 (39/40) −0.01 (−0.45, 0.43) 0 0.95
  Northern Europe 3 243 (114/129) −0.19 (−0.45, 0.06) 0 0.14
  Southern Europe 1 28 (13/15) 0.23 (−0.052, 0.97) — 0.55
  Middle East/Asia 0 — — — —
  Oceania 1 176 (58/118) −0.08 (−0.38, 0.21) — 0.57
 Fat content
  Low fat 4 212 (99/113) 0.01 (−0.26, 0.28) 0 0.93
  High fat 1 91 (13/78) 0.35 (−0.24, 0.94) — 0.24
 Fermentation
  Fermented 4 188 (63/125) 0.10 (−0.23, 0.43) 0 0.55
  Nonfermented 3 174 (79/95) −0.06 (−0.37, 0.24) 0 0.67
 Consistency of dairy products
  Solid 2 131 (33/98) 0.24 (−0.19, 0.66) 0 0.28
  Liquid 6 269 (129/140) −0.01 (−0.26, 0.23) 0 0.91
 Glucose metabolism as primary outcome
  Primary outcome 2 126 (61/65) 0.06 (−0.29, 0.41) 0 0.75
  Not a primary outcome 11 714 (290/424) −0.09 (−0.32, 0.13) 46 0.40
 Type of dietary intervention
  Controlled diet 2 163 (48/114) 0.25 (−0.11, 0.62) 0 0.17
  Free-living (noncontrolled diet) 11 678 (303/375) −0.13 (−0.34, 0.07) 36 0.21
HbA1c Participant BMI
  BMI ≥30 2 112 (57/55) −0.04 (−0.25, 0.17) 0 0.71
  BMI <30 2 400 (200/200) −0.10 (−0.17, −0.03) 0 0.005
 Dairy exposure combined with energy restriction
  No energy restriction 3 474 (237/237) −0.10 (−0.17, −0.03) 0 0.004
  Energy restriction 1 38 (20/18) 0.00 (−0.28, 0.28) — 1.00
 Duration of exposure
  <24 wk 2 238 (118/120) −0.08 (−0.20, 0.03) 0 0.16
  ≥24 wk 2 474 (139/135) −0.10 (−0.18, −0.02) 0 0.02
 Comorbid conditions
  No metabolic disorder 1 200 (98/102) −0.10 (−0.23, 0.03) — 0.12
  Metabolic disorders 3 312 (159/153) −0.09 (−0.17, −0.01) 0 0.02
 Funding
  Industry 3 474 (237/237) −0.10 (−0.17, −0.03) 0 0.004
  Independent 1 38 (20/18) 0.00 (−0.28, 0.28) — 1.00
 Risk of bias
  Unclear/low risk 2 238 (122/116) −0.09 (−0.17, −0.01) 0 0.02
  Elevated risk 2 274 (135/139) −0.10 (−0.22, 0.02) 0 0.10
Post hoc subgroup analyses for HbA1c Continent
  North America 2 238 (118/120) −0.08 (−0.20, 0.03) 0 0.16
  South America 0 — — — —
  Northern Europe 1 74 (37/37) −0.10 (−0.44, 0.24) — 0.56
  Southern Europe 0 — — — —
  Middle East/Asia 1 200 (102/98) −0.10 (−0.18, −0.02) — 0.02
  Oceania 0 — — — —
 Fat content
  Low fat 2 238 (118/120) −0.08 (−0.20, 0.03) 0 0.16
  High fat 0 — — — —
 Fermentation
  Fermented 0 — — — —
  Nonfermented 1 200 (98/102) −0.10 (−0.23, 0.03) — 0.12
 Consistency of dairy products
  Solid 0 — — — —
  Liquid 3 438 (220/218) −0.09 (−0.16, −0.03) 0 0.006
 Glucose metabolism as primary outcome
  Primary outcome 0 — — — —
  Not a primary outcome 4 512 (257/255) −0.09 (−0.16, −0.03) 0 0.005
 Type of dietary intervention
  Controlled diet 0 — — — —
  Free-living (noncontrolled diet) 4 512 (257/255) −0.09 (−0.16, −0.03) 0 0.005
Open in a new tab

1HbA1c, glycated hemoglobin; T2D, type 2 diabetes.

Fasting blood insulin

Elevated dairy product intake (3.1 servings/d, on average) did not affect fasting blood insulin compared with a minimal dairy intake (0.7 serving/d, on average) (29 studies, n = 1902; MD: −2.97 pmol/L; 95% CI: −7.05, 1.10 pmol/L; P = 0.15, I2 = 21%) (Figure 3). Subgroup analyses suggest an inverse association between fasting insulin and elevated dairy product intake for long-term exposures (P = 0.03) (Table 3). The overall risk of bias was considered “high.” From the studies excluded from the meta-analysis, 2 studies (57, 61) out of 7 observed decreased concentrations of fasting insulin with elevated dairy intake (36, 39, 47, 51, 55, 57, 61).

FIGURE 3.

FIGURE 3

Open in a new tab

Forest plot of the meta-analysis on the effect of increased dairy product intake on fasting blood insulin levels. Weighted mean difference and corresponding 95% CI in fasting insulin levels after elevated dairy product intake in comparison with a minimal intake in randomized controlled trials. Weighted mean differences were calculated with the inverse-variance method in a random-effect model. Each study is represented by gray squares, and the whiskers represent the 95% CI. The black diamond represents the overall mean difference, the center of which corresponds to the pooled weight difference and the extremities represent 95% CI. Statistical heterogeneity is represented by the I2 statistic.

Insulin resistance with the HOMA-IR

Thirteen studies assessed the HOMA-IR index and were pooled in the meta-analysis (Figure 4). A calculation mistake in the HOMA-IR was suspected in 1 study (69), and therefore the SMD was used as a summary measure. No association was observed between elevated dairy intake (2.7 servings/d, on average) and the HOMA-IR index compared with a minimal intake (0.3 servings/d, on average) (13 studies, n = 840; SMD: −0.07; 95% CI: −0.26, 0.12; P = 0.49, I2 = 38%). Subgroup analysis showed no further association (Table 3). The overall risk of bias was considered “high.” From the studies excluded from the meta-analysis, 1 study (57) out of 5 observed a decreased HOMA-IR index after elevated dairy product consumption (36, 38, 43, 51, 57).

FIGURE 4.

FIGURE 4

Open in a new tab

Forest plot of the meta-analysis on the effect of increased dairy product intake on HOMA-IR. Values are weighted standardized mean differences and corresponding 95% CIs in HOMA-IR after elevated dairy product intake in comparison with a minimal intake in randomized controlled trials. Weighted standardized mean differences were calculated with the inverse-variance method in a random-effects model. Each study is represented by gray squares, and the whiskers represent the 95% CIs. The black diamond represents the overall standardized mean difference, the center of which corresponds to the pooled weight differences, and the extremities represent 95% CI. Statistical heterogeneity is represented by the I2 statistic.

HbA1c

HbA1c was inversely associated with elevated dairy product intake (2.7 servings/d, on average) in comparison with a minimal intake (0.8 servings/d, on average) (4 studies, n = 512; MD: −0.09%; 95% CI: −0.16%, −0.03%; P = 0.005, I2 = 0%) (Figure 5) (34, 49, 51, 53, 64, 69). The following subgroup analyses suggest a reduction in HbA1c with elevated dairy intake (Table 3): mean BMI <30 (P = 0.005), metabolic disorders (P = 0.02), studies funded by industry (P = 0.004), exposure order 24 wk (P = 0.02), no energy restriction (P = 0.04), unclear/low risk of bias (P = 0.02), studies conducted in Asian countries (P = 0.02), liquid dairy products (P = 0.006), free-living trials (P = 0.005), and studies in which glucose metabolism was not the primary outcome (P = 0.005). The overall risk of bias was considered “unclear.”

FIGURE 5.

FIGURE 5

Open in a new tab

Forest plot of the meta-analysis on the effect of increased dairy product intake on glycated hemoglobin. Values are weighted mean differences and corresponding 95% CIs in glycated hemoglobin concentrations after elevated dairy product intake in comparison with a minimal intake in randomized controlled trials. Weighted mean differences were calculated with the inverse-variance method in a random-effects model. Each study is represented by gray squares, and the whiskers represent the 95% CIs. The black diamond represents the overall mean difference, the center of which corresponds to the pooled weight differences, and the extremities represent 95% CI. Statistical heterogeneity is represented by the I2 statistic.

Meta-regressions were performed a posteriori in order to address the association between increasing servings of dairy products and T2D-related parameters. No dose-response relation was observed for fasting blood glucose, fasting blood insulin, HOMA-IR, and HbA1c (data not shown).

Publication bias and quality of evidence

Funnel plots for all outcomes are available online (Supplemental Figures 2–5). Publication bias for HbA1c could not be detected from funnel plots due to the small sample sizes. Fasting glucose and fasting insulin showed unusual funnel plots, where most studies were grouped at the top of the graphs. No publication bias was detected for HOMA-IR index. The overall quality of evidence was very low for fasting blood glucose, fasting blood insulin, and HbA1c. The quality of evidence was considered low for HOMA-IR index and HbA1c (Table 2).

 

 

Discussion

In this systematic review, a significant increase in fasting blood glucose was observed with elevated dairy intake, whereas fasting blood insulin and HOMA-IR were not significantly changed. However, a negative association between elevated dairy product intake and HbA1c was also observed in 4 studies of ≥12 wk. The overall quality of evidence of the effect of elevated dairy product intake on T2D risk factors was very low to low.

Elevated intake of dairy products was positively associated with fasting glucose concentrations in 33 studies. This result is concordant with a previous meta-analysis of randomized controlled trials, which reported an elevation in fasting glucose concentrations with elevated dairy product intake (18). Nevertheless, the elevation of fasting glucose concentrations associated with elevated dairy product intake remains modest (MD of 0.07 mmol/L); therefore, the clinical significance of the data remains uncertain. It should be noted that glucose concentrations can be modulated by other sources of variation, some of which were taken into account in subgroup analyses. First, weight variation and physical activity may greatly affect fasting glucose. Further, the diet consumed in parallel with dairy products can greatly affect glucose concentrations, especially in studies with free-living designs. Without strict control of the diet, extrinsic factors can modulate the effects of the main dietary intervention, with subconscious compensatory behaviors, for example. Further, studies with lower risk of bias were associated with increased fasting glucose concentrations in comparison with studies with a high risk of bias, suggesting that study methodology could have had a significant influence on results. Differences were also observed between geographic locations, suggesting differences between genetic background and intervariability in the response of elevated dairy product intake on fasting glucose concentrations.

Fasting insulin concentrations were not associated with high dairy product intake in 29 studies. However, a modest significant reduction in fasting insulin (MD of 5.4 pmol/L) was associated with elevated dairy intake in long-term intervention studies. A reduction in insulinemia and increased fasting glucose concentrations could suggest a reduction in pancreatic β cell function and further progression of T2D, but these results should be interpreted with caution due to the small sample size and the high risk of bias of the studies. Further, the wide ranges of variation measures in the meta-analysis suggest important interindividual variability within studies or limited power due to small sample sizes. In addition, the reduction in fasting insulin concentrations in long-term studies remains small; therefore, the change in fasting insulin concentrations might not be sufficient to have clinical significance.

High dairy product intake was not associated with changes in the HOMA-IR in 13 studies (SMD of −0.07). Similarly, a meta-analysis of randomized controlled trials observed no beneficial effect of low-fat dairy products in comparison with whole-fat dairy intake on insulin resistance based on HOMA-IR (18). Another systematic review of clinical trials assessed the effect of increased dairy product intake on glucose metabolism and insulin sensitivity and suggested that elevated dairy intake could potentially improve insulin sensitivity (19). However, to confirm the potential beneficial effects of dairy products on insulin sensitivity, the authors underlined the need for larger and longer studies with subjects at risk of developing T2D (19).

Finally, the meta-analysis noted a negative association between high dairy product intake and HbA1c in 4 studies of ≥12 wk. The inclusion of HbA1c as an outcome gives a better overview of the potential risk of T2D since this parameter is often used as a main parameter for T2D diagnosis (74). The results of the present systematic review suggest that increasing dairy intake might be beneficial to improve HbA1c (MD: −0.09%); however, the reduction observed remains modest. Improvement in HbA1c is concordant with evidence from cohort studies, many of which suggesting a beneficial role of dairy product intake on the incidence of T2D (12, 14, 15). Despite the relevant information given by HbA1c about the progression of T2D in clinical trials, the incidence of T2D as a primary outcome was not addressed in the included studies. Clinical trials with nutrition interventions are often short-term interventions with a limited number of subjects, which greatly limits the possibility of observing new T2D diagnosis (75). Furthermore, the number of included studies assessing HbA1c was limited in the present meta-analysis, which prevents firm conclusions about the inverse association observed between elevated dairy product intake and HbA1c. Although most studies assessing HbA1c had a moderate risk of bias compared with most studies in the systematic review, the quality of evidence remains low due to the indirectness of evaluating HbA1c as a primary outcome. Thus, suggesting modifications to the actual recommendations for dairy product intake for T2D prevention is highly limited based on the available evidence.

The subgroup analyses performed in the systematic review suggest that many cofounding factors might influence the effect of elevated dairy intake on T2D risk factors, such as length of exposure, the risk of bias, the country of origin, the consistency of the dairy product administered, or BMI. Surprisingly, increased intake of fermented dairy products did not improve any of T2D-related biochemical parameters in the present systematic review. These results contrast with evidence from meta-analysis of cohort studies suggesting an inverse association between cheese and the risk of developing T2D, and a nonlinear association between yogurt intake and the risk of T2D (12). Given that yogurt is often seen as a marker of a healthy diet, the association observed in cohort studies might be influenced by other environmental factors such as other healthy foods. The differences between population characteristics and serving sizes in the included studies might have been too heterogeneous to observe the beneficial effects expected from fermented dairy products. Further, subgroup analyses for fermented dairy products combined yogurt and cheese in the same subgroup; however, yogurt and cheese have different food matrixes and variable amounts of fat content, which are likely to affect T2D risk factors in very specific ways (76). In sum, the subgroup analyses presented can be used to explore new avenues to better understand the role of increased dairy product intake in T2D risk factors.

Regarding the limitations of the systematic review, the interpretation of results is limited by the overall high risk of bias of the included studies. Further, most interventions were conducted in free-living conditions; consequently, dietary elements or food combinations could have interfered with dairy product intake. Further, there were no criteria for types and validity of dietary data collection tools and compliance assessment in the present review, which are often major sources of errors in nutritional clinical trials (75). No information on nutritional biomarkers of dairy products was collected, but it should be noted that very few studies use biomarkers as a compliance tool. These missing elements may have introduced further bias in the meta-analysis (75). Finally, all subjects without T2D, with or without hyperglycemia or insulin resistance, could have underestimated the association due to a dilution of the effect. This limitation was partially resolved with subgroup analysis by observing metabolic comorbid conditions. This systematic review also has several strengths, such as the exhaustive and inclusive search strategy. Finally, the use of GRADE approach for assessing quality of evidence is another strength of the present review (30).

Conclusions

Overall, elevated dairy intake was associated with increased fasting glucose concentrations; however, the clinical significance of this remains uncertain. Fasting blood insulin and HOMA-IR were not associated with dairy product intake. Elevated dairy intake was associated with a modest reduction in HbA1c levels, but the small number of included studies prevent firm conclusions about the clinical significance of this result being drawn. In addition, due to the high risk of bias of most included studies and the very low quality of evidence for fasting glucose and fasting insulin and HbA1c, results are mixed and should be interpreted with caution. Rigorous long-term and adequately powered trials are required. Moreover, the incorporation of genetic evaluation and nutritional biomarkers would be interesting for future studies.

Supplementary Material

Supplemental File
Click here for additional data file. (75.2KB, pdf)

Acknowledgments

We thank Alexis Turgeon and Kaoutar Ennour-Idrissi for their valuable advice given throughout the protocol elaboration, the review process, analysis, and for the revision of the manuscript. The authors’ responsibilities were as follows—SO: designed the research, analyzed the data and performed the statistical analysis, and wrote the first draft of the manuscript; SO and A-FT: conducted the research; A-FT, CG, and IR: critically reviewed the manuscript; IR: had primary responsibility for final content; and all authors: read and approved the final manuscript.

Notes

SO received a scholarship from the Canadian Institutes of Health Research (CIHR); CG holds a Junior 2 Research Scholar from the Fonds de Recherche du Québec–Santé (FRQ-S); IR holds a Junior 1 Research Scholar from the FRQ-S.

Author disclosures: SO, A-FT, CG, and IR, no conflicts of interest.

Supplemental Material 1 and Supplemental Figures 1–5 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/advances/.

Abbreviations used; HbA1c, glycated hemoglobin; MD, mean difference; SMD, standardized mean difference; T2D, type 2 diabetes.

References

  • 1. World Health Organization Global report on diabetes [Internet]. Geneva: World Health Organization; 2017. Available from: http://www.who.int/diabetes/global-report/en/. [Google Scholar]
  • 2. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998;15:539–53. [DOI] [PubMed] [Google Scholar]
  • 3. Seuring T, Archangelidi O, Suhrcke M. The economic costs of type 2 diabetes: a global systematic review. Pharmacoeconomics 2015;33:811–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Chen L, Magliano DJ, Zimmet PZ. The worldwide epidemiology of type 2 diabetes mellitus—present and future perspectives. Nat Rev Endocrinol 2011;8:228–36. [DOI] [PubMed] [Google Scholar]
  • 5. Punthakee Z, Goldenberg R, Katz P. Diabetes Canada 2018 clinical practice guidelines for the prevention and management of diabetes in Canada: definition, classification and diagnosis of diabetes, prediabetes and metabolic syndrome. Can J Diabetes 2018;42:S10–5. [DOI] [PubMed] [Google Scholar]
  • 6. International Diabetes Federation DF diabetes atlas [Internet]. 8th ed Brussels: International Diabetes Federation; 2017. Available from: http://www.diabetesatlas.org. [Google Scholar]
  • 7. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–9. [DOI] [PubMed] [Google Scholar]
  • 8. Geloneze B, Vasques ACJ, Stabe CFC, Pareja JC, Rosado LEFP de L, Queiroz EC de, Tambascia MA. HOMA1-IR and HOMA2-IR indexes in identifying insulin resistance and metabolic syndrome: Brazilian Metabolic Syndrome Study (BRAMS). Arq Bras Endocrinol Metabol 2009;53:281–7. [DOI] [PubMed] [Google Scholar]
  • 9. Chartrand D, Da Silva MS, Julien P, Rudkowska I. Influence of amino acids in dairy products on glucose homeostasis: the clinical evidence. Can J Diabetes 2017;41:329–37. [DOI] [PubMed] [Google Scholar]
  • 10. Mozaffarian D, Cao H, King IB, Lemaitre RN, Song X, Siscovick DS, Hotamisligil GS. Trans-palmitoleic acid, metabolic risk factors, and new-onset diabetes in US adults. Ann Int Med 2010;153:790–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Candido FG, Ton WTS, Alfenas R de CG. Dairy products consumption versus type 2 diabetes prevention and treatment; a review of recent findings from human studies. Nutr Hosp 2013;28:1384–95. [DOI] [PubMed] [Google Scholar]
  • 12. Aune D, Norat T, Romundstad P, Vatten LJ. Dairy products and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis of cohort studies. Am J Clin Nutr 2013;98:1066–83. [DOI] [PubMed] [Google Scholar]
  • 13. Tong X, Dong J-Y, Wu Z-W, Li W, Qin L-Q. Dairy consumption and risk of type 2 diabetes mellitus: a meta-analysis of cohort studies. Eur J Clin Nutr 2011;65:1027–31. [DOI] [PubMed] [Google Scholar]
  • 14. Gao D, Ning N, Wang C, Wang Y, Li Q, Meng Z, Liu Y, Li Q. Dairy products consumption and risk of type 2 diabetes: systematic review and dose-response meta-analysis. PLoS One 2013;8:e73965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Gijsbers L, Ding EL, Malik VS, de Goede J, Geleijnse JM, Soedamah-Muthu SS. Consumption of dairy foods and diabetes incidence: a dose-response meta-analysis of observational studies. Am J Clin Nutr 2016;103:1111–24. [DOI] [PubMed] [Google Scholar]
  • 16. Internal Clinical Guidelines Team Type 2 diabetes in adults: management. London: National Institute for Health and Care Excellence (UK); 2015. [PubMed] [Google Scholar]
  • 17. Sievenpiper JL, Chan CB, Dworatzek PD, Freeze C, Williams SL. Nutrition therapy. Can J Diabetes 42:S64–79. [DOI] [PubMed] [Google Scholar]
  • 18. Benatar JR, Sidhu K, Stewart RAH. Effects of high and low fat dairy food on cardio-metabolic risk factors: a meta-analysis of randomized studies. PLoS One 2013;8:e76480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Turner KM, Keogh JB, Clifton PM. Dairy consumption and insulin sensitivity: a systematic review of short- and long-term intervention studies. Nutr Metab Cardiovasc Dis 2015;25:3–8. [DOI] [PubMed] [Google Scholar]
  • 20. Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions, version 5.1.0 [updated March 2011]. The Cochrane Collaboration ; 2011. Available fromwww.handbook.cochrane.org. [Google Scholar]
  • 21. O'Connor S, Turcotte AF, Taschereau-Charron A, Gagnon C, Rudkowska I. Comparison between high and low dairy product intake in the prevention of type 2 diabetes: an updated systematic review. PROSPERO: international prospective register of systematic reviews. Year of registration 2017. CRD42017080137 Available from: http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42017080137. [Google Scholar]
  • 22. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009;151:264–9, W64. [DOI] [PubMed] [Google Scholar]
  • 23. Wong SS-L, Wilczynski NL, Haynes RB. Developing optimal search strategies for detecting clinically sound treatment studies in EMBASE. J Med Libr Assoc 2006;94:41–7. [PMC free article] [PubMed] [Google Scholar]
  • 24. Wong SSL, Wilczynski NL, Haynes RB. Optimal CINAHL search strategies for identifying therapy studies and review articles. J Nurs Scholarsh 2006;38:194–9. [DOI] [PubMed] [Google Scholar]
  • 25. The Cochrane Collaboration Data collection form, intervention review—RCTs [Internet]. Cochrane Airways. Available from: http://airways.cochrane.org/resources-and-information. [Google Scholar]
  • 26. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [DOI] [PubMed] [Google Scholar]
  • 27. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Health Canada Eating well with Canada's food guide. HC Pub 4651. Ottawa: Queen's Printer; 2007. [Google Scholar]
  • 29. Review Manager (RevMan) [Computer program] Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration; 2014.
  • 30. Berkman ND, Lohr KN, Ansari M, McDonagh M, Balk E, Whitlock E, Reston J, Bass E, Butler M, Gartlehner G et al.. Grading the strength of a body of evidence when assessing health care interventions for the effective health care program of the agency for healthcare research and quality: an update. Methods guide for effectiveness and comparative effectiveness reviews [Internet]. Rockville (MD): Agency for Healthcare Research and Quality; 2008[cited 2017 Dec 13]. Available from: http://www.ncbi.nlm.nih.gov/books/NBK174881/. [PubMed] [Google Scholar]
  • 31. Alemán-Mateo H, Macías L, Esparza-Romero J, Astiazaran-García H, Blancas AL. Physiological effects beyond the significant gain in muscle mass in sarcopenic elderly men: evidence from a randomized clinical trial using a protein-rich food. Clin Interv Aging 2012;7:225–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Alemán-Mateo H, Carreón VR, Macías L, Astiazaran-García H, Gallegos-Aguilar AC, Enríquez JRR. Nutrient-rich dairy proteins improve appendicular skeletal muscle mass and physical performance, and attenuate the loss of muscle strength in older men and women subjects: a single-blind randomized clinical trial. Clin Interv Aging 2014;9:1517–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Arnberg K, Mølgaard C, Michaelsen KF, Jensen SM, Trolle E, Larnkjær A. Skim milk, whey, and casein increase body weight and whey and casein increase the plasma C-peptide concentration in overweight adolescents. J Nutr 2012;142:2083–90. [DOI] [PubMed] [Google Scholar]
  • 34. Barr SI, McCarron DA, Heaney RP, Dawson-Hughes B, Berga SL, Stern JS, Oparil S. Effects of increased consumption of fluid milk on energy and nutrient intake, body weight, and cardiovascular risk factors in healthy older adults. J Am Diet Assoc 2000;100:810–7. [DOI] [PubMed] [Google Scholar]
  • 35. Bendtsen LQ, Blaedel T, Holm JB, Lorenzen JK, Mark AB, Kiilerich P, Kristiansen K, Astrup A, Larsen LH. High intake of dairy during energy restriction does not affect energy balance or the intestinal microflora compared with low dairy intake in overweight individuals in a randomized controlled trial. Appl Physiol Nutr Metab 2018;43:1–10. [DOI] [PubMed] [Google Scholar]
  • 36. Bernini LJ, Simão ANC, Alfieri DF, Lozovoy MAB, Mari NL, de Souza CHB, Dichi I, Costa GN. Beneficial effects of Bifidobacterium lactis on lipid profile and cytokines in patients with metabolic syndrome: a randomized trial. Effects of probiotics on metabolic syndrome. Nutrition 2016;32:716–9. [DOI] [PubMed] [Google Scholar]
  • 37. Bowen J, Noakes M, Clifton PM. Effect of calcium and dairy foods in high protein, energy-restricted diets on weight loss and metabolic parameters in overweight adults. Int J Obes (Lond) 2005;29:957–65. [DOI] [PubMed] [Google Scholar]
  • 38. Brassard D, Tessier-Grenier M, Allaire J, Rajendiran E, She Y, Ramprasath V, Gigleux I, Talbot D, Levy E, Tremblay A et al.. Comparison of the impact of SFAs from cheese and butter on cardiometabolic risk factors: a randomized controlled trial. Am J Clin Nutr 2017;105:800–9. [DOI] [PubMed] [Google Scholar]
  • 39. Buchowski MS, Aslam M, Dossett C, Dorminy C, Choi L, Acra S. Effect of dairy and non-dairy calcium on fecal fat excretion in lactose digester and maldigester obese adults. Int J Obes (Lond) 2010;34:127–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Crichton GE, C Howe PR, Buckley JD, Coates AM, Murphy KJ. Dairy consumption and cardiometabolic health: outcomes of a 12-month crossover trial. Nutr Metab 2012;9:19–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Dönmez N, Kısadere İ, Balaban C, Kadiralieva N. Effects of traditional homemade koumiss on some hematological and biochemical characteristics in sedentary men exposed to exercise. Biotech Histochem 2014;89:558–63. [DOI] [PubMed] [Google Scholar]
  • 42. Drouin-Chartier J-P, Gagnon J, Labonté M-È, Desroches S, Charest A, Grenier G, Dodin S, Lemieux S, Couture P, Lamarche B. Impact of milk consumption on cardiometabolic risk in postmenopausal women with abdominal obesity. Nutr J 2015;14:12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Farnsworth E, Luscombe ND, Noakes M, Wittert G, Argyiou E, Clifton PM. Effect of a high-protein, energy-restricted diet on body composition, glycemic control, and lipid concentrations in overweight and obese hyperinsulinemic men and women. Am J Clin Nutr 2003;78:31–9. [DOI] [PubMed] [Google Scholar]
  • 44. Ghadirian P, Shatenstein B, Verdy M, Hamet P. The influence of dairy products on plasma uric acid in women. Eur J Epidemiol 1995;11:275–81. [DOI] [PubMed] [Google Scholar]
  • 45. Gilbert J-A, Joanisse DR, Chaput J-P, Miegueu P, Cianflone K, Alméras N, Tremblay A. Milk supplementation facilitates appetite control in obese women during weight loss: a randomised, single-blind, placebo-controlled trial. Br J Nutr 2011;105:133–43. [DOI] [PubMed] [Google Scholar]
  • 46. Hulston CJ, Churnside AA, Venables MC. Probiotic supplementation prevents high-fat, overfeeding-induced insulin resistance in human subjects. Br J Nutr 2015;113:596–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Jacobsen R, Lorenzen JK, Toubro S, Krog-Mikkelsen I, Astrup A. Effect of short-term high dietary calcium intake on 24-h energy expenditure, fat oxidation, and fecal fat excretion. Int J Obes (Lond) 2005;29:292–301. [DOI] [PubMed] [Google Scholar]
  • 48. Benatar JR, Jones E, White H, Stewart RAH. A randomized trial evaluating the effects of change in dairy food consumption on cardio-metabolic risk factors. Eur J Prev Cardiol 2014;21:1376–86. [DOI] [PubMed] [Google Scholar]
  • 49. Jones KW, Eller LK, Parnell JA, Doyle-Baker PK, Edwards AL, Reimer RA. Effect of a dairy- and calcium-rich diet on weight loss and appetite during energy restriction in overweight and obese adults: a randomized trial. Eur J Clin Nutr 2013;67:371–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Josse AR, Atkinson SA, Tarnopolsky MA, Phillips SM. Increased consumption of dairy foods and protein during diet- and exercise-induced weight loss promotes fat mass loss and lean mass gain in overweight and obese premenopausal women. J Nutr 2011;141:1626–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51. Lee YJ, Seo JA, Yoon T, Seo I, Lee JH, Im D, Lee JH, Bahn K-N, Ham HS, Jeong SA et al.. Effects of low-fat milk consumption on metabolic and atherogenic biomarkers in Korean adults with the metabolic syndrome: a randomised controlled trial. J Hum Nutr Diet 2016;29:477–86. [DOI] [PubMed] [Google Scholar]
  • 52. Lee Y, Ba Z, Roberts RF, Rogers CJ, Fleming JA, Meng H, Furumoto EJ, Kris-Etherton PM. Effects of Bifidobacterium animalis subsp. lactis BB-12® on the lipid/lipoprotein profile and short chain fatty acids in healthy young adults: a randomized controlled trial. Nutr J 2017;16:39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53. Machin DR, Park W, Alkatan M, Mouton M, Tanaka H. Hypotensive effects of solitary addition of conventional nonfat dairy products to the routine diet: a randomized controlled trial. Am J Clin Nutr 2014;100:80–7. [DOI] [PubMed] [Google Scholar]
  • 54. Maersk M, Belza A, Stødkilde-Jørgensen H, Ringgaard S, Chabanova E, Thomsen H, Pedersen SB, Astrup A, Richelsen B. Sucrose-sweetened beverages increase fat storage in the liver, muscle, and visceral fat depot: a 6-mo randomized intervention study. Am J Clin Nutr 2012;95:283–9. [DOI] [PubMed] [Google Scholar]
  • 55. Melanson EL, Donahoo WT, Dong F, Ida T, Zemel MB. Effect of low- and high-calcium dairy-based diets on macronutrient oxidation in humans. Obes Res 2005;13:2102–12. [DOI] [PubMed] [Google Scholar]
  • 56. Pu F, Guo Y, Li M, Zhu H, Wang S, Shen X, He M, Huang C, He F. Yogurt supplemented with probiotics can protect the healthy elderly from respiratory infections: a randomized controlled open-label trial. Clin Interv Aging 2017;12:1223–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57. Rideout TC, Marinangeli CPF, Martin H, Browne RW, Rempel CB. Consumption of low-fat dairy foods for 6 months improves insulin resistance without adversely affecting lipids or bodyweight in healthy adults: a randomized free-living cross-over study. Nutr J 2013;12:56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58. Rosado JL, Garcia OP, Ronquillo D, Hervert-Hernández D, Caamaño MDC, Martínez G, Gutiérrez J, García S. Intake of milk with added micronutrients increases the effectiveness of an energy-restricted diet to reduce body weight: a randomized controlled clinical trial in Mexican women. J Am Diet Assoc 2011;111:1507–16. [DOI] [PubMed] [Google Scholar]
  • 59. Shlisky JD, Durward CM, Zack MK, Gugger CK, Campbell JK, Nickols-Richardson SM. An energy-reduced dietary pattern, including moderate protein and increased nonfat dairy intake combined with walking promotes beneficial body composition and metabolic changes in women with excess adiposity: a randomized comparative trial. Food Sci Nutr 2015;3:376–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60. Smilowitz JT, Wiest MM, Teegarden D, Zemel MB, German JB, Van Loan MD. Dietary fat and not calcium supplementation or dairy product consumption is associated with changes in anthropometrics during a randomized, placebo-controlled energy-restriction trial. Nutr Metab (Lond) 2011;8:67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61. Soerensen KV, Thorning TK, Astrup A, Kristensen M, Lorenzen JK. Effect of dairy calcium from cheese and milk on fecal fat excretion, blood lipids, and appetite in young men. Am J Clin Nutr 2014;99:984–91. [DOI] [PubMed] [Google Scholar]
  • 62. Stancliffe RA, Thorpe T, Zemel MB. Dairy attentuates oxidative and inflammatory stress in metabolic syndrome. Am J Clin Nutr 2011;94:422–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63. St-Onge M-P, Goree LLT, Gower B. High-milk supplementation with healthy diet counseling does not affect weight loss but ameliorates insulin action compared with low-milk supplementation in overweight children. J Nutr 2009;139:933–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64. Tanaka S, Uenishi K, Ishida H, Takami Y, Hosoi T, Kadowaki T, Orimo H, Ohashi Y. A randomized intervention trial of 24-wk dairy consumption on waist circumference, blood pressure, and fasting blood sugar and lipids in Japanese men with metabolic syndrome. J Nutr Sci Vitaminol 2014;60:305–12. [DOI] [PubMed] [Google Scholar]
  • 65. Thompson WG, Rostad Holdman N, Janzow DJ, Slezak JM, Morris KL, Zemel MB. Effect of energy-reduced diets high in dairy products and fiber on weight loss in obese adults. Obes Res 2005;13:1344–53. [DOI] [PubMed] [Google Scholar]
  • 66. Torres MRSG, Francischetti EA, Genelhu V, Sanjuliani AF. Effect of a high-calcium energy-reduced diet on abdominal obesity and cardiometabolic risk factors in obese Brazilian subjects. Int J Clin Pract 2010;64:1076–83. [DOI] [PubMed] [Google Scholar]
  • 67. Tripolt NJ, Leber B, Blattl D, Eder M, Wonisch W, Scharnagl H, Stojakovic T, Obermayer-Pietsch B, Wascher TC, Pieber TR et al.. Short communication: Effect of supplementation with Lactobacillus casei Shirota on insulin sensitivity, β-cell function, and markers of endothelial function and inflammation in subjects with metabolic syndrome—a pilot study. J Dairy Sci 96:89–95. [DOI] [PubMed] [Google Scholar]
  • 68. Van Loan MD, Keim NL, Adams SH, Souza E, Woodhouse LR, Thomas A, Witbracht M, Gertz ER, Piccolo B, Bremer AA et al.. Dairy foods in a moderate energy restricted diet do not enhance central fat, weight, and intra-abdominal adipose tissue losses nor reduce adipocyte size or inflammatory markers in overweight and obese adults: a controlled feeding study. J Obes 2011;2011:989657. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69. Wennersberg MH, Smedman A, Turpeinen AM, Retterstøl K, Tengblad S, Lipre E, Aro A, Mutanen P, Seljeflot I, Basu S et al.. Dairy products and metabolic effects in overweight men and women: results from a 6-mo intervention study. Am J Clin Nutr 2009;90:960–8. [DOI] [PubMed] [Google Scholar]
  • 70. Zemel MB, Thompson W, Milstead A, Morris K, Campbell P. Calcium and dairy acceleration of weight and fat loss during energy restriction in obese adults. Obes Res 2004;12:582–90. [DOI] [PubMed] [Google Scholar]
  • 71. Zemel MB, Richards J, Milstead A, Campbell P. Effects of calcium and dairy on body composition and weight loss in African-American adults. Obes Res 2005;13:1218–25. [DOI] [PubMed] [Google Scholar]
  • 72. Zemel MB, Donnelly JE, Smith BK, Sullivan DK, Richards J, Morgan-Hanusa D, Mayo MS, Sun X, Cook-Wiens G, Bailey BW et al.. Effects of dairy intake on weight maintenance. Nutr Metab (Lond) 2008;5:28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73. Zemel MB, Teegarden D, Loan MV, Schoeller DA, Matkovic V, Lyle RM, Craig BA. Dairy-rich diets augment fat loss on an energy-restricted diet: a multicenter trial. Nutrients 2009;1:83–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74. Canadian Diabetes Association Clinical Practice Guidelines Expert Committee , Goldenberg R, Punthakee Z. Definition, classification and diagnosis of diabetes, prediabetes and metabolic syndrome. Can J Diabetes 2013;37(Suppl 1):S8–11. [DOI] [PubMed] [Google Scholar]
  • 75. Gibson TM, Ferrucci LM, Tangrea JA, Schatzkin A. Epidemiological and clinical studies of nutrition. Semin Oncol 2010;37:282–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76. Panahi S, Tremblay A. The potential role of yogurt in weight management and prevention of type 2 diabetes. J Am Coll Nutr 2016;35:717–31. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental File
Click here for additional data file. (75.2KB, pdf)

Articles from Advances in Nutrition are provided here courtesy of American Society for Nutrition

RESOURCES