Table 2.

Risk of Bias table (Based on QUIPS^a)

Author	1. Study participation		2. Study attrition		3. Determinant measurement			4. Outcome measurement		5. Study confounding										6. Statistical analysis and reporting
										5a. Confounders measured						5d. Confounders accounted for in analysis
	1a. Consecutive series of participants	1b. Adequate participation rate (> 70%)	2a. Adequate follow-up rate (≥80%)	2b. No important differences between participants and drop-out	3a. ≥70% complete data for each determinant	3b. Method and setting of the measurement is the same for all study participants	3c. Appropriate methods of imputation	4a. Outcome measurement truly captures volunteering	4b. Method and setting of measurement is the same for all study participants	5a1. Age	5a2. Socioeconomic Status	5a3. Gender	5a4. Participation in voluntary work at baseline	5b. Method and setting of measurement is the same for all study participants	5c. Appropriate methods of imputation	5d1. Age	5d2. Socioeconomic Status	5d3. Gender	5d4. Participation in voluntary work at baseline	6a. Statistical model adequate for study design	6b. No overfitting	6c. No selective reporting of results
Ajrouch et al. [18]	+	+	–	?	+	+	N.A.b	+	+	+	+	+	–	+	N.A.	+	+	+	–	+	+	+
Bartels et al. [19]	+	?	?	?	?	+	?	+	+	–	+	+	+	+	?	–	+	+	+	+	+	+
Bekkers [20]	–	?	–	+	?	+	?	+	+	+	+	+	+	+	?	–	–	–	–	+	+	+
Broese van Groenou & Van Tilburg [21]	+	–	–	?	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+	+
Choi & Chou [22]	+	–	–	+	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+/−c	+
Cramm & Nieboer [23]	–	–	–	–	?	+	N.A.	+	+	+	+	+	+	+	N.A.	+	+	+	+	+	–	+
Curl et al. [24]	+	?	?	?	+	+	N.A.	+	+	+	+	+	+	+	N.A.	+	+	+	+	+	+	+
Curl et al. [25]	+	?	?	?	+	+	N.A.	+	+	+	+	+	+	+	N.A.	+	+	+	+	+	+	+
Einolf & Philbrick [26]	+	?	+	?	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+	+
Hank & Erlinghagen [27]	+	–	–	?	?	?	?	+	?	+	+	+	+	?	?	+	+	+	+	+	+	+
Johnston [28]	+	–	–	?	?	+	?	+	+	+	+	+	+	+	?	–	+	+	+	+	+	+
Lim & Mac Gregor [29]	+	?	–	?	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+	+
McNamara & Gonzales [30]	+	?	?	?	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Mike et al. [31]	+	?	?	?	+	+	N.A.	+	+	+	+	+	+	+	N.A.	+	+	+	+	+	+	+
Nesbit [32]	+	?	?	?	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+	+
Okun et al. [33]	+	?	?	?	+	+	N.A.	+	+	+	+	+	+	+	N.A.	+	+	+	+	+	+	+
Parkinson [34]	+	?	–	?	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+	+
Pavlova & Silbereisen [35]	+	+/−d	?e	−/+f	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Pavlova & Silbereisen [36]	+	–	–	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Son & Wilson [37]	+	+	–	?	?	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Son & Wilson [38]	+	+	–	?	?	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Son & Wilson [39]	+	+	–	?	?	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Voorpostel & Coffé [40]	+	?	–	?	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+	+
Voorpostel & Coffé [41]	+	?	–	?	?	+	?	+	+	+	+	+	+	+	?	+	+	+	+	+	+	+

^aQUIPS Quality of Prognosis Studies in Systematic Reviews. Assessment: + (Yes) (represents low risk of bias); - (No) (represents high risk of bias); ? (Unclear) (represents uncertain risk of bias, insufficient information was available to assess the risk of bias)

^bNot Applicable

^cFor the outcome volunteer engagement (starting) there is no over fitting, so low risk of bias, but for the outcome volunteer cessation (quitting), there is slight over fitting of the model, so high risk of bias

^dBaseline participation in the first sample (age group 16-43) was adequate (77%), but the baseline participation in the second sample (age group 56-75) not (52,9%)

^eNo information is provided on the follow-up rates. However, the second sample (age group 56-75) is the same as the sample used in Pavlova et al. 2016 and attrition is higher than 20%

^fAttrition in the first sample (age group 16-43) was selective w.r.t. volunteering at T1, for the second sample (age group 56-75) attrition was not selective w.r.t. volunteering at T1