Impact of Diet and Exercise on Weight and Cognition in Older Adults: a rapid review

Abstract

Objective:

To determine where the current literature stands in regards to diet/exercise interventions on cognition in overweight or obese individuals.

Data Source:

A rapid review was conducted of English-language studies published in Medline from January 1965 to January 2020.

Study Inclusion and Exclusion Criteria:

Included studies were intervention studies lasting ≥12 weeks, with participants aged ≥65 years, with a body mass index ≥25kg/m²

Data Extraction:

Data extracted included study population, duration, intervention design, outcomes, and results.

Data Synthesis:

Outcomes were qualitatively measured due to paucity of RTC.

Results:

1845 citations were identified, 31 full-text articles were reviewed, and 5 studies were included. Studies had usual care control groups and combined exercise/diet intervention groups with 31–3,526 participants randomized to each arm. Mean age of participants was 69.2–83.4 years. Studies reporting on cognitive changes showed marginally significant positive changes in cognition, and those that reported BMI indicated potential improvements in cognition.

Conclusions:

The number of interventions assessing the combined effects of both diet and exercise is low. Future studies should evaluate the impact of combined effects to ascertain whether cognitive decline may be reversed in older adults with a BMI ≥25kg/m².

OBJECTIVE

The aging process increases the probability of chronic diseases such as cardiovascular disease,¹ cancer,² and cognitive impairment.³ While improvements in prevention strategies and advanced cardiac therapies⁴ have led to marked reductions in mortality for both cardiovascular disease and cancer,⁵ the incidence rate of dementia continues to increase rapidly with age,³ as rates double every 5 years from ages 65–90.⁶ Emerging disease-modifying pharmacotherapies are in development,⁷ but existing medications are only marginally effective.⁸ Hence, there is reliance on nonpharmacological therapies such as diet, exercise, and behavioral therapy in managing cognitive decline.

Over the past four decades, rates of obesity have steadily increased to over 40% in older adults,⁹ and the number of obesity-related comorbidities also increase with age.¹⁰ Obesity increases the risk of disability, nursing home placement, and early mortality in older adults.¹¹ In mid-life, classified as having obesity increases one’s risk for developing long-term cognitive impairment¹² and this relationship is observed even when fat is accumulated later in life in older adults.¹³ Despite high quality trials to suggest safety and efficacy of diet and activity interventions in preventing cognitive decline,¹⁴ implementation of these intensive interventions in routine care is hindered by a considerable number of barriers¹⁵ including access, staffing, and funding. In addition, common mechanisms including inflammation, and vascular and metabolic dysfunction ¹⁶ lead to impairment in both physical function¹¹ and cognitive impairment in persons with obesity¹⁶ making participating in exercise programs challenging.

While health promotion interventions consisting of dietary modifications and physical activity have independently demonstrated improvements in cardiometabolic variables, physical function and quality of life in older adults,^15,17 it is unclear whether combined diet and activity interventions can effectively mitigate the development of new cognitive impairment, or promote reversal or stabilization of impairments in those with known cognitive dysfunction.^18–20 Previous systematic reviews have demonstrated the effectiveness of diet-only or exercise-only interventions in older adults (with or without impaired cognition) for supporting long-term cognition.^21–23 However, there is a dearth of studies evaluating the impact of combined diet/exercise interventions in this population. Our aim was to conduct a rapid review to evaluate the evidence for multi-component (diet and physical activity) interventions to improve /protect cognition in overweight and obese older adults with or without pre-intervention cognitive impairment.

METHODS

A rapid review, defined as a form of knowledge synthesis in which components of the systematic review process are simplified or omitted to produce information in a timely manner,²⁴ was conducted using the reporting guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).²⁵ Supplementary Table 1 outlines each component of this checklist.

Data Sources

A review was conducted of English-language studies published in Medline (Pubmed) from January 1965 to January 2020. Our date range was intentionally kept broad as no prior rapid or systematic reviews evaluating the impact of combined dietary and exercise interventions in older adults with obesity or overweight, with or without cognitive impairment, on outcomes of cognition were identified. The literature search included subject headings and keywords to capture previously published concept terms related to older adults and obesity, and incorporated cognition as an additional element. Duplicate studies were removed. To facilitate the rapid evaluation of this review, we purposefully limited our search to Medline. See Supplementary Table 2 for our full search strategy.

Inclusion and Exclusion Criteria

Our initial intent was to focus on randomized controlled trials; however, given the paucity of RCTs on this subject, we chose to include pre/post quasi-experimental and non-randomized trials as well. Letters, case reports, editorials, and other gray literature (websites, conference proceedings, abstract submissions, clinical trial registries) were excluded. As we identified systematic reviews, study personnel reviewed the relevant bibliographies for additional citations (MNR, VKR). We included studies with community-dwelling older adults with a mean age ≥ 65 years who were classified as overweight or obese (body mass index ≥25kg/m² and ≥30kg/m², respectively), and interventions lasting a minimum of 12 weeks. Studies were excluded if they enrolled participants with end-stage dementia. To broaden our reach, we included studies describing any kind of intentional multi-component weight-loss intervention consisting of any dietary modification and activity intervention. All studies were required to have a minimum duration of 12 weeks. Outcome measures included any cognitive or functional measures, or weight/BMI outcomes. Studies were excluded if they were not published in the English language, non-human, or if they included participants <60 years old. Any studies involving the use of medications, non-protein dietary supplements alone (e.g., without an exercise component), or surgical procedures as part of the intervention were excluded. Inter-reviewer reliability was assessed with a small portion of the search results prior to conducting the full title/abstract review, with ≥ 80% concordance between reviewers required. Two reviewers (MNR, VKR) manually reviewed title/abstracts for inclusion/exclusion criteria, and a third reviewer (LMS) reconciled any discrepancies prior to full-text review.

Data Extraction

The search results were initially uploaded into Rayyan, an online web-application that assists in identifying relevant citations in the conduct of a systematic review (https://https://rayyan.qcri.org/). Similar to systematic reviews, we used the PICO framework (Patients, Intervention, Controls, Outcomes) to evaluate studies meeting criteria.

Tables 2–4 incorporate all the data abstracted from the full-text articles. Our primary purpose was to evaluate multicomponent (diet and activity) health promotion interventions in older adults with and without cognitive dysfunction. As such, we abstracted information about study design including year of the study; country; number of arms and number of participants, the duration of the study, descriptions of cognitive status, description of the intervention groups, and the study setting. Information on the study population was also collected including age, age-range, proportion of females and ethnicity/race, comorbidities, baseline/follow up (and change in) weight, body mass index, waist circumference, functional status were abstracted. Change in weight, physical function and quality of life and indications of statistical significance were also abstracted.

Table #2:

Description of the Study Population of the Randomized Controlled Trials

Reference	Year	Country	Cognitive Status	Selection Criteria	Arms	# per arm	Study Duration
Cameron30	2013	Australia	Normal cognition	Age ≥70 years; ≥ 3 CHS frailty criteria; not living in residential care facility; residing locally; MMSE >18; not an ongoing client of DRACS; life expectancy >12 months; not a research participant in another physical intervention.	Control: health and age-related services normally available to older people	121	12 months
					Intervention: multifactorial, interdisciplinary program targeting frailty individually tailored to: nutritional intake; psychiatrists/ psychologists referral; home-based physiotherapy sessions	120
Ngandu 32	2015	Finland	↑ risk for dementia; normal cognition to MCI	Age 60–77 years; CAIDE risk score ≥6, word list memory task result ≤19 words; CERAD word list recall ≤75% or MMSE 20–26. Exclusion: diagnosed/suspected dementia; malignant disease, major depression; symptomatic CV disease; revascularization within 1 year; impaired vision, hearing or ability to communicate.	Control: regular health advice	599	24 months
					Intervention: multidomain lifestyle intervention program focused on nutrition, exercise, cognitive training and management of vascular risk factors	591
Rosenburg41	2018	Finland	↑ dementia risk, normal to MCI	Age 60–77. CAIDE risk score of ≥6; word list memory task result ≤ 19 words; CERAD word list recall ≤ 75% or MMSE 20–26. Exclusion: diagnosed dementia, suspected dementia at screening visit, conditions affecting participation in the interventions (impaired vision, hearing or ability to communicate)	Control: general health advice	591	24 months
					Intervention: multidomain lifestyle intervention program focused on nutrition, exercise, cognitive training and management of vascular risk factors	599
Van De	2014	Netherlands	Normal to	Age ≥65 years; frailty or pre-frailty. Exclusion:	Control: no protein, no exercise	31	24 weeks
Rest42			MCI	cancer, COPD, muscle disease, type 2 DM, renal insufficiency, silent ischemia	Intervention: Resistance exercise: warm up on cycle ergometer, leg press/extension, chest press, lateral pull-down, pec-dec, vertical row. Placebo supplement: no protein, 7.1 g lactose, 0.4 g calcium, consumed after breakfast/lunch	31
					Intervention: Protein: twice daily, 250 ml protein beverage supplement (15 g protein, 7.1 g lactose, 0.5 g fat, 0.4 g calcium), consumed after breakfast/lunch; no exercise	34
					Intervention: Combination of protein + exercise above	31
Van de Middelaar43	2018	Netherlands	Normal cognition	Age 70–78 years. Exclusion: dementia, any condition likely to hinder long-term follow-up (terminal illness, alcoholism).	Control: standard care according to prevailing standards for cardiovascular risk management	3526	6–8 years
					Intervention: 4 monthly RN visits giving individually tailored lifestyle advice on smoking, diet, physical activity, weight and blood pressure.

Abbreviations: CAIDE- Cardiovascular Risk Factors, Aging, and Incidence of Dementia; CERAD-Consortium to Establish a Registry for Alzheimer’s Disease; CHS – Cardiovascular Health Study; COPD- Chronic obstructive pulmonary disease; CV- Cardiovascular; DM- Diabetes Mellitus; DRACS- Division of Rehabilitation and Aged Care Services; MCI – Mild Cognitive Impairment; MMSE – Mini-Mental Status Examination; RN – registered nurse.

Table #4:

Study Outcomes of Randomized Controlled Trials

Reference	Intervention	Baseline Weight or BMI	Δ Weight or BMI	Δ Weight or BMI, p-value	Change in Function	Inter-group Δ Function	Change in Cognitive Domains	Inter-group Δ Cognition
Cameron30	Control: health and age-related services normally available to older people	69.3±17.3 kg	1.48±5.09 kg	Δ= 0.78 (−0.72, 2.28); p= 0.31	Gait Speed: 0.0019±0.230 m/s Chair stands: 0.31±0.98	Gait speed: 0.068 (0.012,0.123); p=0.02 Chair Stands 0.33 (0.07,0.59); p=0.01	NR	NR
	Intervention: multifactorial, interdisciplinary program targeting frailty and individually tailored to each participant based on: nutritional intake; psychiatrists/ psychologists referral; home-based physiotherapy sessions	68.5±15.7 kg	0.72±5.98 kg		Gait Speed: 0.049±0.183 m/s Chair stands: - 0.14±1.22		NR	NR
Ngandu 32	Control: regular health advice	28.1±4.9 kg/m2	−0.33±0.05b kg/m2	−0.077 kg/m2 (−0.149,−0.006); p=0.02	−0.22±0.05ab	0.033 (−0.035,0.101); p=0.34	NTB: 0.16±0.01b Executive: 0.06±0.02b Processing: 0.04±0.02b Memory: 0.31±0.02b Abbr. Memory: 0.17±0.03b	NTB: 0.022 (0.002,0.42) Executive: 0.027 (0.001,0.052) Processing 0.030 (0.003,0.057) Memory: 0.015 (−0.017,0.048) Abbr.Memory: 0.038 (0.002,0.73)
	Intervention: multidomain lifestyle intervention program focused on nutrition, exercise, cognitive training and management of vascular risk factors	28.3±4.5 kg/m2	−0.49±0.05 kg/m2		−0.16±0.05 ab		NTB: 0.20±0.01b Executive: 0.11±0.02 b Processing: 0.10±0.10 b Memory: 0.34±0.02b Abbr.Memory: 0.24±0.03 b
Rosenburg 41	Control - general health advice	a28.1±4. 9 kg/m2	NR	NR	NR	NR	BMI<27.4kg/m2: 0.038 (0.010,0.067) BMI>27.4kg/m2: 0.004 (−0.024,0.032)	P=0.28d
	Intervention: Multidomain lifestyle intervention program nutrition, exercise, cognitive training and management of vascular risk factors.	28.3±4.5 kg/m2	NR		NR
Van de Rest 42	Control - no protein, no exercise	26.6±3.2 kg/m2	NR	NR	NR	NR	Episodic: 0.20± 0.79 Attention: −0.12± 0.69 Processing: 0.13±0.50 Executive: 0.07±0.42	Episodic: p=0.51 Attention: p=0.02 Processing: p=0.32 Executive: P=0.78
	Intervention: Resistance exercise: warm up on cycle ergometer, leg press/extension, chest press, lateral pull-down, pec-dec, vertical row. Placebo supplement: no protein, 7.1 g lactose, 0.4 g calcium, consumed after breakfast/lunch	28.2±4.6 kg/m2	NR		NR		Episodic: −0.02±0.57 Attention: −0.35±0.70 Processing: 0.07±0.48 Executive: −0.02±0.42
	Intervention: Protein: twice daily, 250 ml protein beverage supplement (15 g protein, 7.1 g lactose, 0.5 g fat, 0.4 g calcium), consumed after breakfast/lunch; no exercise	27.0±4.6 kg/m2	NR		NR		Episodic: 0.01±0.57 Attention: 0.04± 0.57 Processing: −0.23±0.19 Executive: 0.17± 0.43	Episodic: p=0.79 Attention: p=0.38 Processing: p=0.04 Executive: p=0.09
	Intervention: Combination of protein + exercise above	28.7±4.5 kg/m2	NR		NR		Episodic: 0.07±0.62 Attention: 0.19±0.63 Processing: 0.08±0.51 Execulimitive: 0.04±0.44
Van Middelaar 43	Low LIBRAc	25.9±3.1 kg/m2	NR	NR	NR	NR	All-cause dementia: Ix vs. Control: 5.7 vs. 6.4%	0.88 (0.54,1.43)
	Intermediate LIBRAc	26.7±3.6 kg/m2	NR		NR		All-cause dementia: Ix vs. Control: 6.2 vs. 6.7%	0.91 (0.57,1.47)
	High LIBRAc	29.6±4.6 kg/m2	NR		NR		All-cause dementia: Ix vs. Control: 7.4 vs. 7.9%	0.92 (0.59,1.41)

Abbreviations: Abbrev – Abbreviated; BMI – body mass index; Ix- intervention; LIBRA – Lifestyle for Brain Health index; NR – not reported; NTB – neuropsychological test battery represents a composite score of 14 other cognitive tests; Δ – difference between groups

^anumbers represent differences in short performance physical battery, a measure of physical performance ²⁸

^bNumbers reflect standard error

^cLow, intermediate and high according to the LIBRA index which is a Dementia risk algorithm.

^dp-value for interaction between BMI category and difference between intervention and control group on primary outcome of change in neuropsychological test battery

Data Synthesis

Our outcomes were qualitatively evaluated as significant heterogeneity precluded meta-analysis. To evaluate the quality of the included studies included in the review, we applied the Cochrane Collaboration’s Risk-of-Bias Tool²⁶ (Table 1).This tool permits the assessment of whether studies have inherent biases in their design, allowing investigators to evaluate their quality. We evaluated the following components: sequence generation; allocation concealment; blinding of participants, data and outcome assessors; and whether there was incomplete or selective outcome reporting. Three reviewers conducted this assessment (MNR VKR, LMS), rating each component as high, low, or unclear risk of bias for each criterion. The senior author (JAB) adjudicated discrepancies in risk-of-bias assessments.

Table #1 –

Methodological Quality of Randomized Controlled Studies (n=5) - Cochrane Risk-of-Bias Tool

Reference	Year of Study	Overall Riskb	Sequence Generation	Allocation Concealment	Blinding			Incomplete Outcome Data	Selective Outcome Reporting	Other Sources of Bias
					Participants	Data Collectors	Outcome Assessors
Cameron 30	2013	Unclear	Low	Low	Unclear	Unclear	Unclear	Low	Low	Unclear
Ngandu 32	2015	Unclear	Low	Low	Low	Low	Low	Low	Low	Unclear
Rosenburg 41	2018	Unclear	Low	Low	Low	Low	Low	Low	Low	Unclear
Van De Rest 42	2014	Low	Low	Low	Low	Low	Low	Low	Low	Low
Van Middelaar 43	2018	Unclear	Low	Low	Unclear	Unclear	Unclear	Low	Low	Unclear
	# Low	1	5	5	3	3	3	5	5
	% Lowa	20%	100%	100%	60%	60%	60%	100%	100%

^aProportion of full-text articles that fulfill low degree of bias

^bOverall Risk of Bias is represented as “Low” if all components are listed as Low (e.g., high quality), “Unclear” if any component is listed as “Unclear” or “High” if any of components are listed as “High”

RESULTS

Figure 1 outlines the study flow of the 1,813 citations identified. We evaluated an additional 32 studies from systematic reviews. Of the 1,845 citations reviewed, 31 articles warranted full-text review (using Endnote X8, Thomson Reuters), five articles were finally included that were based on unique study populations. Our complete search strategy is outlined in Supplementary Table 2.

Figure 1:

Flow diagram of study selection process for the review including the systematic review bibliographies. These articles were accounted for in the flow diagram. RCT indicates randomized controlled trial.

Risk of Bias Assessment

Table 1 describes the bias assessment of the five included studies using the Cochrane Collaboration’s Risk-of-Bias Tool. The evaluation was subjectively conducted by two independent reviewers (VKR, MNR), who agreed on all components. Overall risk of bias designation was considered low (e.g., high quality study) only if all components of the assessment tool were considered low – only one of the studies fulfilled this criterion. However, of the five included studies three fulfilled criteria of low bias in each of the individual categories.

Study Characteristics

All but one of the included studies were conducted in Europe, and none in the United States (Table 2).The number randomized in each arm ranged from 31 to 3,526, with a median length of follow-up of 2 years. Selection criteria varied across all the studies. Importantly, the interventions were heterogeneous and consisted of both nutrition and exercise-based components, ranging from counseling or nutritional supplementation to active treatment strategies that varied (Tables 2 and 3). The studies included in this review were led by dietitians (n=3) or nurses (n=1), or participants were provided protein supplementation (n=1). Control groups offered usual care consisting of general counseling. All studies were conducted in community-based older adults not residing in a care facility. Race was not reported by the included studies.

Table #3:

Baseline Characteristics of Randomized Controlled Trials

Reference	Arm	Age ± S.D. (years)	Female Sex (%)	% White (Culture)	Baseline Functional Status	BMI (kg/m2)
Cameron30	Control: health and age-related services normally available to older people	83.2±5.9	82 (67.7)	NR	Barthel Index Scoree 92.5±14.3	26.4±6.04
	Intervention: multifactorial, interdisciplinary treatment program targeting frailty individually tailored to specific criteria: nutritional intake; referrals to psychiatrists/ psychologists; home-based physiotherapy sessions	83.4±5.8	81 (66.9)	NR	Barthel Index Scoree 93.9±11.1	26.1±5.91
Ngandu 32	Control: regular health advice	69.2±4.7	284 (47.4)	NR (Finish)	Physical Activity two or more times per week-72%a	28.1±4.9
	Intervention: multidomain lifestyle intervention program focused on nutrition, exercise, cognitive training and management of vascular risk factors	69.5±4.6	267 (45.2)	NR (Finish)	Physical Activity two or more times per week-70%	28.3±4.5
Rosenburg41	Control - general health advice	69.2±4.7	284 (47.4)	NR (Finish)	1.4 (0.3–8.9)	28.1±4.9
	Intervention: multidomain lifestyle intervention program focused on nutrition, exercise, cognitive training and management of vascular risk factors	69.5±4.6	267 (45.2)	NR (Finish)	1.4 (0.3–8.2)	28.3±4.5
Van de Rest42	Control - no protein, no exercise	81.2±7.4	16 (52)	NR	Frailty scored: 1.6±0.8	26.2±3.2
	Intervention: Resistance exercise: warm up on cycle ergometer, leg press/extension, chest press, lateral pulldown, pec-dec, vertical row. Placebo supplement: no protein, 7.1 g lactose, 0.4 g calcium, consumed after breakfast/lunch	79.2±6.3	21 (68)	NR	Frailty scored: 2.0±1.3	28.2±4.6
	Intervention: Protein: twice daily, 250 ml protein beverage supplement (15 g protein, 7.1 g lactose, 0.5 g fat, 0.4 g calcium), consumed after breakfast/lunch; no exercise	77.9±8.1	20 (59)	NR	Frailty scored: 1.7±1.2	27.0±4.6
	Intervention: Combination of protein + exercise above	77.7±8.8	20 (64)	NR	Frailty scored: 1.5±0.7	28.7±4.5
Van de Middelaar43	Low LIBRAb	74.2±2.5	563 (51.6)	1057 (96.9)	1065 (97.6%)c	25.9
	Intermediate LIBRAb	74.3±2.5	562 (51.9)	1042 (96.4)	990 (91.6%)c	26.7
	High LIBRAb	74.5±2.5	657 (59.6)	1054 (95.6)	784 (71.1%)c	29.6

Abbreviations: BMI – body mass index; LIBRA – Lifestyle for Brain Health index; MMSE – mini-mental status examination; NR – not reported; SD – standard deviation

^aThis study outlined self-reported medical disorders data for this is not shown. Baseline SPPB was not presented in the manuscript

^bLow, intermediate and high according to the LIBRA index, is a Dementia risk algorithm ⁴⁴.

^crepresented as number of physically active participants in each group (percentage of physically active)

^dFrailty score – score determining frailty where three or more are present: unintentional weight loss, self-reported exhaustion, weakness, slow walking speed, and low physical activity. A higher score indicates increasingly frail;

^eBarthel Index Score (range 0–100) a higher score indicates greater independence.

Participant Characteristics

All studies reported study populations with characteristics of a mean participant body mass index (BMI) in the overweight range (≥ 25kg/m²), but none had a mean BMI classified as having obesity (≥ 30kg/m²). Mean age of study samples ranged from 69.2 to 83.4 years and most study populations included predominantly female. Functional status was defined differently, using Barthel Index,²⁷ short performance physical battery,²⁸ or physical activity and frailty score.²⁹ Participants’ cognitive status was classified as normal to mild cognitive impairment; each study used different definitions to define this measure.

Interventions and Outcomes

Intervention types and reporting of data were markedly heterogeneous across the five included studies (see Table 4 for full descriptions of interventions and outcomes per study), which included the types of nutrition and exercise interventions delivered as part of the study. The study conducted by Cameron³⁰ and colleagues evaluated a multifactorial, interdisciplinary program including individually tailored diet and exercise components designed to target frailty. This was the only study to evaluate changes in weight. While this group reported non-significant changes in weight and quality of life, as represented by the EuroQol-5D³¹ visual analog scale, they did not evaluate cognition. Ngandu³² reported a significant change in BMI between groups (−0.077kg/m² [−0.149,−0.006]; p=0.02), that was not associated with any significant changes in short performance physical battery, as the result of a multidomain lifestyle program. The remaining studies focused on changes in cognition and demonstrated marginally significant changes in overall neuropsychological functioning, executive function that controls higher level thoughts and actions, and processing speed. None of the studies evaluated weight-loss or change in physical function as a primary outcome. It was clear, though, that in the five studies whose mean BMI was classified as overweight, there were signals of changes in cognition despite inconsistent intervention delivery across studies.

DISCUSSION

Our results highlight a lack of empirical research involving well-designed randomized controlled trials evaluating the effects of combined dietary and exercise interventions in community-dwelling overweight or obese older adults, on important geriatric-specific outcomes such as weight and cognition. The inconsistencies and dearth of data provide an opportunity to address future gaps in research.

This rapid review highlights a critical gap in health promotion research exploring the link between overweight/obesity, diet and physical activity interventions, and cognition, among older adults in community-based settings. While the literature suggests the interplay and possible causal relationship between obesity and cognitive decline,³¹ there is a major gap in combined dietary and exercise interventions aimed at their treatment and mitigation. Combined interventions appeared to improve physical function, as previously described.¹⁷ However, the impact of combined interventions on cognition has not previously been systematically evaluated. Despite heterogeneous assessment of cognitive outcomes, this review provides some preliminary data to suggest some improvements with multi-component lifestyle interventions. Importantly, elements of dietary intervention, whether protein supplementation, caloric restriction, or other nutritional supplementation, coupled with aerobic and resistance exercises should be tested for the effects on cognition in future trials.

A wealth of evidence suggests the importance of physical activity in enhancing physical function and reducing the progression of cognitive impairment.^33–35 Home-based regimens in younger participants with subjective memory concerns demonstrated improved cognitive scores, while older adults in the LIFE study only demonstrated improved cognition in those with diabetes.³⁴ Mixed aerobic/resistance exercise show promise in improving both function and cognition.³⁵ Of the five studies meeting our inclusion criteria, four involved in-person exercise interventions led by trained professionals. Supervised exercise regimens enhance compliance and progression that ultimately can improve health outcomes. In addition, specific diet-only trials have demonstrated improved cognition with the Mediterranean diet.³⁶ Only one study evaluated the impact of this diet’s effect on cognition, with or without weight-loss. This is the first RCT to our knowledge that promotes weight loss in addition to participant adherence to the Mediterranean diet that improves cognition.³⁷ While evidence suggests diet interventions are effective, they are multifaceted on their own right and require research on both caloric restriction and nutrient modification in order to develop a standardized practice. The heterogeneity of the dietary interventions suggests the importance of standardized dietary practices and interventions in high-risk populations that either have mild cognitive impairment or may be at risk for functional decline. More rigorous evaluation of dietary interventions combined with exercise interventions are critically needed.

The results of our review prevent us from making any definitive conclusions on whether health promotion interventions could lead to weight loss and improved cognition in older adults with obesity or overweight, with or without cognitive impairment. It is unclear whether the lack of health promotion studies is due to the obesity paradox,³⁸ which necessitates careful consideration of the benefits and risks of weight loss in older adults with obesity,³⁹ or whether excess weight is considered protective in older adults with cognitive issues and so not an area worth targeting for research. While the trials identified in our review were few, we noted that specific elements of bias were consistently classified as having a low risk, suggesting that the majority of the components embedded within these trials were conducted in a rigorous manner. Future randomized controlled trials should consider modelling their design with elements from the authors reviewed in this paper.

This review has a number of limitations. First, our search was limited to the literature indexed by Pubmed and did not include investigation of other established databases, trial registries, or grey literature. While our inclusion criteria did establish that pre/post quasi-experimental and non-randomized trials would be included due to the lack of literature we were unable to find any to include in this review. Second, we focused on community-dwelling older adults; the impact of our findings may not necessarily be applicable to nursing home populations. Third, we deliberately did not include studies with durations that involved less than 12 weeks of intervention, as previous health promotion work suggests that this is the minimum time needed to improve elements of health.⁴⁰ Fourth, heterogeneity in interventions and outcomes of both the interventions and outcome measures made it difficult to make significant interpretations of the data and prevented a formal meta-analysis. Fifth, our findings may be prone to publication bias as negative studies are harder to publish. Lastly, our study sample size is small and lacks diversity which prevents the generalizability of our findings. The impact of race/ethnicity and other potential co-variates importantly may affect outcomes, yet the population was homogeneous, and studies reported important variables erratically.

The public health crisis in the United States and other countries facing Alzheimer’s and related dementias will have lasting physical, social and economic consequences on patients, families, and societies. Both dietary and exercise interventions, separately, have shown efficacy. Our findings were surprising in that we observed a low number of studies, a poor design, and lack of intervention intensity and detail in studies that assess the impact on obesity and cognition. Our findings highlight the need for well-designed, comprehensive evaluation on the effects of multicomponent health promotion interventions on cognitive functioning and cognitive impairment in older adults with obesity or overweight. Importantly, such health promotion interventions need to be scalable for potential widespread dissemination in order to put these research programs into everyday practice.

CONCLUSION

Health promotion interventions that impact cognitive outcomes in older adult with obesity or overweight exist, however, there are few adequate clinical trials that address this population and have both diet and exercises interventions. Future trials should focus on enhancing dietary quality and exercise engagement in high-risk populations as a possible means to ameliorate decline cognition and function, along with weight loss.

Supplementary Material

1

Supplementary Table 1: Components of the reporting items for the Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Supplementary Table 2: Search Criteria from Medline

SO WHAT?

What is already known on this topic?

Current research has investigated interventions consisting of dietary modifications and physical activity independently and has demonstrated improvements in cardiometabolic variables, physical function and quality of life in older adults.^15,17 However, it is unclear what the effects are of a combined diet and activity interventions.

What does this article add?

This article highlights the critical gap in knowledge in regards to interventions that combine both diet and exercise in overweight and obese older adults.

What are the implications for health promotion practice or research?

Future interventions should focus on engaging a high-risk population in a dual combination therapy of consisting of both diet and exercise to help prevent further cognitive decline. As our population grows older and trends in obesity continue to rise it is critical to have research programs that can be integrated easily into clinical practice.