Table 2 Associations of clustering patterns of diet, PA and sedentary behaviours with adiposity in children, adolescents, and young adults (n = 16)
Authors (Year) | Clusters summary | Health outcomes | Method of analysis | Covariates | Sex-stratified associations | Results |
|---|---|---|---|---|---|---|
Children (n = 3) | ||||||
Bel-Serrat et al., (2019) [65] | C1. Physically active and healthy diet ↑VPA ↑FV ↓CSD ↓ST C2. Healthy diet ↑FV ↓CSD *C1 and C2 were observed in all four groups C3. Physically active ↑VPA * Cluster 3 was observed in East Europe, South Europe/Mediterranean countries and West-Central Asia C4. Physically active and sedentary ↑VPA ↑ST * Cluster 4 only emerged in the North European countries C5. Sedentary and physically inactive ↓VPA ↑ST * North Europe, South Europe/Mediterranean countries and West-Central Asia C6. Low beverage intake, low sedentary and physically inactive ↓VPA ↓FV ↓CSD ↓ST *Cluster 6 was present in all the groups except in North Europe C7. High beverage intake and F&V intake ↑FV ↑CSD * Cluster 7 was present in both North Europe and West-Central Asia C8. Sedentary, physically inactive and healthy diet ↓VPA ↑FV ↓CSD ↑ST *Cluster 8 was emerged in North European countries C9. High beverage intake, sedentary and physically inactive ↓VPA ↑CSD ↑ST *Cluster 9 comprised in the North European and East European countries C10. Sedentary and physically active ↑VPA ↑ST *Cluster 10 was observed in East Europe only C11. High beverage intake, sedentary and physically active ↑VPA ↑CSD ↑ST *Cluster 11 was emerged in both East Europe and in South Europe/Mediterranean countries C12. Sedentary, physically active and healthy diet ↑VPA ↑FV ↓CSD ↑ST * Cluster 12 was specific to South Europe/Mediterranean countries C13. Physically active, high beverage intake, sedentary and high F&V intake ↑VPA ↑FV ↑CSD ↑ST *Cluster 13 was only observed in West-Central Asia | BMI/A z-scores Weight status: • Underweight/healthy weight • Overweight/obese | Mixed-effects regression | Sex, age, parental education level and season of completion of the questionnaire | No | South Europe/Mediterranean All Cs except C3  + BMI/A  + overweight/obese East Europe C2, C6, C9, and C10 + BMI/A C2, C6, C9, and C10 + overweight/obese North Europe C8 + BMI/A C8 + overweight/obese C2, C4, and C5 + overweight/obese |
Gubbels et al., (2012) [70] | C1. Sedentary-snacking pattern ↑UHF ↑TV C2. Healthy intake pattern ↑HF C3. Sandwich pattern ↑UHF ↑HF C4. Sporty-traditional meal pattern ↑PA ↑HF | BMI z-score Weight status: • Overweight | Backward regression | Child sex, BMI z-score at age 5 years, general appetite and activity style; parental educational level, working hours, country of birth and BMI | No | C1 + BMI at age 7 years and 8 years C1 + overweight at age 7 years |
Santaliestra-PasÃas et al., (2015) [67] | C1 ↑PA ↓FV ↓SSB ↓SB C 2 ↓PA ↓FV ↓SSB ↑SB C 3 ↑PA ↓FV ↓SSB ↑SB C 4 ↓PA ↓FV ↑SSB ↓SB C 5 ↓PA ↓FV ↓SSB ↓SB C 6 ↓PA ↑FV ↓SSB ↓SB | BMI z-score Waist circumference z-score Sum of skinfolds z-score | ANCOVA and logistic regression | SES and age | Yes | C2 and C3 + BMIz, WCz, and SSz B in C2 + BMIz and WCz greater than one |
Adolescents (n = 9) | ||||||
Cuenca-GarcÃa et al., (2013) [63] | C1. Healthy diet and active ↑MVPA ↑DQ ↓ST ~ SBHW C2. Healthy diet and academic  ~ MVPA ↑DQ ~ ST ↑SBHW C3. Healthy diet and inactive ↓MVPA ~ DQ ~ ST ~ SBHW C4. Unhealthy diet and screen user  ~ MVPA ↓DQ ↑ST ~ SBHW C5. Unhealthy diet and active ↑MVPA ↓DQ ~ ST ~ SBHW | Body fat percentage Fat-free mass percentage Waist circumference | ANOVA | –- | Yes | 0 |
Cureau et al., (2018) [55] | C1 ↓MVPA ↓FIB ↑ST | Weight status: • Overweight/obesity Abdominal obesity | Poisson regression | Brazilian regions, sex, age categories, skin colour, economic index, and school type | Yes | C1 + overweight/obesity and abdominal obesity |
Dantas et al., (2018) [51] | C 1 ↑PA ↓FV ↓SSB ↓SB C 2 ↓PA ↓FV ↓SSB ↑SB C 3 ↑PA ↓FV ↓SSB ↑SB C 4 ↓PA ↓FV ↑SSB ↓SB C 5 ↓PA ↓FV ↓SSB ↓SB C 6 ↓PA ↑FV ↓SSB ↓SB | Weight status: • Overweight and obesity | Binary logistic regression | Age and economy class | Yes | B in C2 63% ↑ chance to have overweight and obesity G in C2 53% ↑ chance to have overweight and obesity B in C4 51% ↑ chance to have overweight and obesity G in C4 47% ↑ chance to have overweight and obesity |
de Mello et al., (2021) [56] | Whole sample C1 ↓PA ↓FV ↑SSS ↑ST C2 ↑PA ↑FV ↓SSS ↑ST Boys C1 ↓PA ↑FV ↑SSS ↑ST C2 ↑PA ↑FV ↓SSS ↑ST Girls C1 ↓PA ↑FV ↑SSS ↑ST C2 ↑PA ↓FV ↑SSS ↑ST | Weight status: • Overweight including obesity • Non-overweight including thinness and normal weight | Logistic regression | Age and maternal education | Yes | 0 |
Moreira et al., (2018) [64] | Among boys, clusters 1 to 3, and 5, were similar in both HELENA and ELANA studies whereas cluster 4 showed different behaviours C1 ↓MVPA ↓FV ↓SSB ↑TV C 2 ↓MVPA ↓FV ↑SSB ~ TV C 3 ↑MVPA ↓FV ↓SSB ↓TV C 4 In the HELENA study, cluster 4 was characterized by: ↓MVPA ↑FV ↓SSB ↓TV In the ELANA study, cluster 4 was characterized by: ↑MVPA ~ FV ↑SSB ↑TV C 5 ↓MVPA ↓FV ↓SSB ↓TV Among girls, clusters showed similarities in both studies C 1 ↓MVPA ↓FV ↓SSB ↑TV C 2 ↓MVPA ↓FV ↑SSB ↓TV C 3 ↑MVPA ↓FV ↓SSB ↓TV C 4 ↓MVPA ↑FV ↓SSB ↓TV whereas in ELANA girls this cluster presented as: ↑MVPA ↑FV ↑SSB ↓TV C 5 ↓MVPA ↓FV ↓SSB ↓TV | BMI z-score: • Overweight (including obesity) Waist circumference z-score Body fat percentage z-score | Logistic regression | Total energy intake in both studies, SES in the HELENA study, and type of school in the ELANA study | Yes | HELENA B, C2 + WC and %BF ELANA B, C1 + WC and %BF ELANA B, C4 + WC ELANA G, C3 and C4 + BMI |
Spengler et al., (2014) [72] | C 1 ↑PA ~ DQ ~ MU C 2 ↓PA ↑DQ ↓MU C 3 ↓PA ↓DQ ↑MU C 4 ↓PA ↓DQ ↓MU | Weight status: • Normal weight • Overweight (including obesity) | Multinomial logistic regression and ANOVA | –- | Yes | Weight status change C2, C3, and C4 increased overweight % from T1 to T2 C3 highest overweight % and greatest increase in T2 G in C2 and C4 increase overweight in T2 B in C3 had significant change in weight status and largest increase of overweight members Older age in C2, C3, and C4 had significant increase in overweight Greatest change in weight status over time between younger and older was in C3 Age and SES were predictors for changing in weight status C3 members were more likely to change from normal weight to overweight over a period of six years |
Van der Sluis et al., (2010) [73] | C1. Healthy ↑PA ↑FV ↓SN ↓CSD ↓SB C2. Quite healthy ↑PA ~ FV ↓SN ↓CSD ~ SB C3. Quite unhealthy  ~ PA ↓FV ↓SN ↓CSD ↑SB C4. Unhealthy ↓PA ↓FV ↑SN ↑CSD ↑SB | BMI (kg/m2) | Linear regression | Sex and parental education level | No | C4 - BMI |
Veloso et al., (2012) [74] | C1. Active gamers ↑PA ↑SW ↑CSD ↓FV ↑ST C2. Healthy group ↑PA ↓SW ↓CSD ↑FV ↓SB C3. Sedentary group ↓PA ~ SW ~ CSD ↓FV ↓SB ~ TV | Weight status: • Normal weight • Overweight • Obese | ANOVA | –- | No | C2 + BMI than C1 C3 + BMI |
Wadolowska et al., (2018) [75] | C1. Prudent-Active ↑VPA ↑HF ↑BF ↓ST C2. Fast-food-Sedentary ↑UHF ↓BF ↑ST C3. notPrudent-notFast-food-lowActive ↓VPA ↓HF ↓UHF ↓BF | Central obesity Weight status: • Overweight/obesity | Logistic regression | Sex, age, residence, family affluence scale, and nutrition knowledge score | No | C1 - central obesity and overweight/obesity (lowest) C2 + central obesity (highest) and overweight/obesity C3 + central obesity and overweight/obesity (highest) In C1, 47% ↓ chance of central obesity and 33% ↓ chance of overweight/obesity than in C3 In C2, 2.22% ↑ chance of central obesity than in C1 |
Children and Adolescents (n = 4) | ||||||
Khoshhali et al., (2021) [90] | C1 ↓FV ↓DP C2 ↑ PA ↑SW ↑SSB ↑SN ↑FF ↓FV ↓DP C3 ↑ PA ↓BF ↓FV ↓DP ↑FF ↑SW ↑SSB ↑SN | Actual and perceived weight status | Multilevel logistic regression | Age, living area (urban vs. rural), child physical activity, number of children in the home, screen time, mother education, father education, parent physical activity, and parent weight status | Yes | Compared to C1, OR of having C2 for G who perceived themselves as overweight/obese were less than those who perceived themselves as normal weight, and G who their parent perceives them as overweight/obese was more than those who their parent perceives them as normal Underweight G were 37% more likely to be in C3 rather than normal-weight G Compared to C1, OR of having C2 for B who their parent perceives them as underweight was more than those who their parent perceives them as normal B who their parent perceives them as overweight/obese were 27% more likely to be in C3 rather than those who their parents perceive them as normal Significantly ↓ scores of UHB for G and B who perceived themselves as overweight/obese |
Leech et al., (2015) [59] | C1. Most healthy ↑MVPA ↓ED ↓SB ↓TV C2. ED consumers who watch TV ↑ED ↑TV C3. High sedentary behaviour/low MVPA ↓MVPA ↑SB | BMI z-score Weight status: • Healthy weight • Overweight/obese | Cross-sectional and longitudinal linear and logistic regression | Sex, age group, maternal education, and clustering by school Longitudinal models were additionally adjusted for baseline BMI Z-score and baseline weight status, respectively | No | Baseline cluster 0 BMI z-score or weight status C2 at baseline ↑odds of overweight/obese at follow-up |
Sánchez-Oliva et al., (2018) [71] | 4 clusters (older children) C1. Healthy lifestyle ↑MVPA ~ MED ↓ST ↓SB C2. Sedentary/healthy diet ↓MVPA ↑MED ↓ST ↑SB C3. High screen  ~ MVPA ↓MED ↑ST ~ SB C4. Low moderate to vigorous physical activity/unhealthy diet ↓MVPA ↓MED ↓ST ↑SB 4 clusters (younger adolescents) C1. Healthy lifestyle ↑MVPA ~ MED ↓ST ↓SB C2. Highly sedentary  ~ MVPA ~ MED ~ ST ↑SB C3. High screen/ unhealthy diet  ~ MVPA ↓MED ↑ST ~ SB C4. Low screen and moderate to vigorous physical activity ↓MVPA ↑MED ↓ST ~ SB 4 clusters (older adolescents) C1. Healthy lifestyle ↑MVPA ↑MED ↓ST ↓SB C2. Sedentary/ healthy diet ↓MVPA ↑MED ↓ST ↑SB C3. High screen  ~ MVPA ~ MED ↑ST ~ SB C4. Low moderate to vigorous physical activity/unhealthy diet ↓MVPA ↓MED ~ ST ~ SB | Body fat percentage | Linear regression | Maternal education, accelerometer wear time, and sex | No | Older children in C1 ↓ BF% at baseline and 2 years later Younger adolescents in C1 ↓ BF% at baseline and 2 years later compared to C3 and C4 Younger adolescents in C1 greater ↓ BF% 2 years later compared to C3 and C4 (BF% changes) Older adolescents in C1 ↓ BF% at baseline compared to C2 BF% at baseline positively predicted BF% 2 years follow-up |
Schroder et al., (2018) [89] | C1. Low physical activity / High screen time / Skipping breakfast / Low meal frequency ↓PA ↓BF ↓MF ↑ST C2. Low physical activity / High screen time / Skipping breakfast ↓PA ↓BF ↑ST C3. Low physical activity / High screen time / Low meal frequency ↓PA ↓MF ↑ST | BMI z-score WHtR Weight status: • Overweight • Obesity Abdominal obesity | General linear models and logistic regression | General linear models adjusted for sex, age, region, community size, maternal education, energy, and energy over and underreporting (BMI z-score and WHtR) Logistic regression adjusted for sex, age, region, community size, maternal education, energy, and energy over and underreporting (overweight and obesity and abdominal obesity) | Yes | A difference of 0.50 SD for BMI z-score and of 0.025 for WHtR between participants with no UHB and those with three or more ↑ UHB ↑ odds of overweight and abdominal obesity 3 UHB ↑ odds of overweight by 168% and ↑ odds of abdominal obesity by 112%, compared to those with none |